Predictors of Employment Outcome for People with Psychiatric Disabilities: A Review of the Literature Since the Mid '80s.

It is generally accepted that work is therapeutic and is an important part of life, filling much of an individual's time, supplying a source of income, providing a source of identity, and contributing to the physiological and psychological well-being in societies (Chan et al., 1997; Dawis, 1987; Mowbray, Bybee, Harris & McCrohan, 1995; Osipow, 1968).

While work serves an important function in the life of a person, impairments in vocational ability, which include choosing, getting, and keeping a job in the community, are considered to be a central feature of mental disorders (Massel et al., 1990). Results of surveys on employment rates of persons discharged from psychiatric hospitals show that the figures for full-time competitive employment range from 20 to 30% (Anthony, Cohen, & Vitals, 1978; Anthony & Jansen, 1984; Dion & Anthony, 1987; Goldstrom & Manderscheid, 1982; Wasylenki, Goering, Lancee, Ballantyne, & Farkas, 1985). For those who are more chronically ill, the figure drops to about 15% (Unger & Anthony, 1984). Clients with severe psychiatric disabilities seem to experience more problems in adjusting to work than clients with other disabilities (McCue & Katz, 1983). Vocational rehabilitation services for psychiatric patients have, therefore, become a focus of concern among mental health professionals. As Bond (1992) pointed out, vocational rehabilitation for persons with severe mental illness was not considered important by rehabilitation centers, even as late as the early 1980's. From the mid 80's onwards, there has been a growing interest in vocational rehabilitation for psychiatric patients rising from the increasing awareness of the low employment rates of people with severe mental illness and the development of supported employment (Mueser et al., 1997). The effort by mental health professionals to design and evaluate vocational rehabilitation programs for patients suffering from psychiatric disabilities is well documented.

With the awareness that employment is a normalizing experience, helping patients with psychiatric disabilities to escape from the role of dependency, "returning to normal functioning" becomes more important than symptom control and reduced hospitalization (Drake, 1998; Mueser et al., 1997). Mental health service providers, therefore, strive to provide consumer-centered services, to offer community-based services, to implement shared decision making, and to enhance quality of life outcomes. All of these efforts make employment the cardinal outcome measure of psychiatric rehabilitation.

Knowing and accepting that employment status, which refers to whether a person is gainfully employed in the labor market, is an important outcome indicator of psychiatric rehabilitation service is not the end of our effort. Two other related questions that have frequently been addressed by both researchers and clinicians are can we predict the employment status of psychiatric patients upon discharge? and if so, how? Identification of predictive variables would tremendously help the psychiatric rehabilitation process to facilitate the employment of their clients. Clinicians could use these predictive variables as a reference when designing programs. Similarly, rehabilitation planners and managers would also gain more information about evaluating the outcome and effectiveness of psychiatric rehabilitation programs.

Early Studies and Reviews

Since the 1960's, there have been numerous studies on the prediction of vocational outcome for individuals with psychiatric disabilities. In a study by Hall, Smith and Shimuknas (1966), it was found that the degree of residual mental illness, marital status, and level of skill were most closely related to the post-hospital employment status of people suffering from mental illness. In another study by Griffiths (1977), work success in the community among people with mental illness is found to be unrelated to intelligence, personality, age and chronicity. A study by Solberg and Chueh (1976) found that participation in occupational therapy programs could predict successful vocational outcome. These are only a few of the studies on the prediction of vocational outcome conducted in the 60's and 70's. There are contradictory findings among such studies. Moreover, many have significant methodological faults. For instance, as pointed out by Massel et al. (1990), diagnosis and psychopathology were often unreliably elicited and recorded; vocational outcome was too vaguely determined. Readers of this literature are easily confused by the inconsistent results and discrepant conclusions of these research projects.

In 1984, Anthony and Jansen conducted a very comprehensive review of the literature on the prediction of the vocational capacity of the chronically mentally ill. Nine major findings were generated from the review of the related literature:

1. Psychiatric symptomatology is a poor predictor of future work performance.

2. Diagnostic category is a poor predictor of future work performance.

3. Intelligence, aptitude, and personality tests are poor predictors of future work performance.

4. A person's ability to function in his or her environment is not predictive of his or her ability to function in a different type of environment.

5. There is little or no correlation between a person's symptomatology and functional skills.

6. The best clinical predictors of future work performance are ratings of a person's work adjustment skills made in a workshop setting or sheltered job site.

7. The demographic predictor of future work performance is the person's prior employment history.

8. A significant predictor of work performance is a person's ability to get along or function socially with others.

9. The best paper-and-pencil test predictors of future vocational performance are tests that measure a person's ego strength or self-concept in the role of worker.

Stauffer (1986) conducted a comprehensive literature review from the 60's to mid 80's. The review centered on six areas: evaluation tools, test scores, demographic data, training programs, staff involvement with clients outside of hospitals, and job and social skills gained by clients in rehabilitation programs. Though some of the results have still not been confirmed by empirical evidence, Stauffer's review concluded that all of these six areas help rehabilitation professionals to predict whether a person with psychiatric disabilities will be able to find and keep a job in the labor market.

Since the mid 80's, especially after the publication of the review article by Anthony and Jansen (1984), research in the field of psychiatric rehabilitation into the prediction of vocational outcome has continued to blossom in the field of psychiatric rehabilitation. Not only have previous findings been substantiated by more sophisticated research methodologies and statistical procedures, but also some new findings have been generated which correct some of the conclusions reviewed above. This article reviews related research efforts since the mid 1980's and summarizes the findings. Recommendations for further research and implications for rehabilitation professionals are discussed.

Method

A literature search was conducted using the CD-ROM service in the Hong Kong Polytechnic University. The databases included Medline, PsycLit, Allied Health and Nursing Abstracts, and Social Work Abstracts. These databases represent collections of articles from the most popular journals and book chapters in the field of psychiatric rehabilitation.

There were several inclusion criteria. One criterion was time frame. Articles published from 1985 to 1997 were included. Also included were articles that reported well-designed research (i.e., with clearly spelled out objectives, subject selection, data collection and data analysis) with people suffering from psychiatric disabilities as the research participants or subjects. Demographic and clinical predictors of employment outcome of people with mental illness should also be reported at some stage of the articles. Key words used in the literature search included prediction, vocation/employment, and mental illness/psychiatric disabilities.

Results

A total of 921 articles (381 from PsycLit, 483 articles from Medline, 43 from Allied Health and Nursing Abstracts, and 14 from Social Work Abstracts) were reviewed. The abstracts of these articles were screened by the principal investigator according to the criteria as stipulated above. Ninety-two articles (30 from PsycLit, 60 from Medline, and two from Allied Health and Nursing Abstracts) were selected for a detailed review.

The 92 articles were separately reviewed by a panel of six specialists. One held the rank of Associate Professor in the field of rehabilitation sciences with the duty of overseeing a Bachelor of Science in Occupational Therapy program. Two were Assistant Professors in rehabilitation sciences specializing in psychosocial rehabilitation. The other three specialists were senior clinicians working in psychosocial settings and with experience in conducting clinical research projects. The review format included the following: author, year, research participants, research design, measures/assessment, results (in terms of significant or non-significant predictors), and limitations. After the intensive review exercises, the results of the reviews were returned to the principal investigator.

Table 1 shows the results of the review in terms of the significant and non-significant predictors. The predictors were then categorized into medical aspects, demographic characteristics, psychosocial aspects, social aspects, functional aspects, environmental aspects and others. The number of studies that cited each of the predictors is shown in Table 2. The results show that premorbid occupational performance, which refers to work skills, work attitude, work adjustment, community living skills and other relevant life experiences before the onset of mental illness, received most support (N=7) in the literature as a significant predictor of post-hospital employment. However, three studies concluded that it was a non-significant predictor. Social skills (N=11) and premorbid functioning (N=6) also received support as a significant predictor of employment outcome from the studies reviewed. Unlike premorbid occupational performance, there were no studies concluding that these latter two aspects were non-significant predictors. The most contradictory aspects were diagnosis and psychiatric symptomatology. Both of these received more or less the same support as a significant predictor (N=9 for symptomatology and N=3 for diagnosis) and as non-significant predictors (N=6 for symptomatology and N=5 for diagnosis). Other aspects that had contradictory results were demographic variables such as age (N=1 as significant predictor and N=2 as non-significant predictor), sex (N=2 as significant predictor and N=5 as non-significant predictor), ethnicity (N=1 as significant predictor and N=2 as non-significant predictor) and marital status (N=2 as significant predictor and N=1 as non-significant predictor). One aspect that also received support as a good predictor was cognitive functioning (N=4); however, there were two studies showing that it was a poor indicator. Other interesting but neglected factors included family relationship and substance abuse. Only one study concluded that family relationship was a good predictor. As to substance abuse, one paper said it was a significant predictor and one study said it was not.

Table 1 Results of review by the panel

 Number
of study    N         Author(s)         Significant predictors

    1      275   Anthony, Rogers,       * Psychiatric
                 Cohen, & Davies          symptomatology
                 (1995)

    2      163   Bailer, J., Brauer,    * Premorbid adjustment
                 W., Rey, E.R. (1996)   * Premorbid social
                                          functioning
                                        * Negative symptoms

    3       79   Beiser, Bean,          * Premorbid work
                 Erickson, Zhang,         performance
                 Iacono, & Rector       * Sex
                 (1994)                 * Putative markers of
                                          biological vulnerability

    4       61   Bell & Lysaker         * Cognitive functioning
                 (1995)                 * Social skills
                                        * Personal presentation
                                        * Negative symptoms

    5       58   Breier, Schreiber,     * Premorbid functioning
                 Dyer & Pickar (1992)

    6       53   Carpenter & Strauss    * Social contact
                 (1991)                 * Stable heterosexual
                                          relationship
                                        * Duration of
                                          hospitalization

    7       46   Charisiou, Jackson,    * Social skills
                 Byole, Burgess,        * Adjustment behaviors
                 Minas, & Joshua
                 (1989)

    8       44   Corrigan, Reedy,       * Negative symptoms
                 Thadani, & Ganet       * QoL
                 (1995)

    9      161   Gaebel & Pietzcker     * Employment duration
                 (1987)                   before admission

   10       75   Gaebel & Pietzcker     * Previous employment
                 (1985)                 * Social functioning
                                        * Use of neuroleptics

   11       44   Geddes, Mercer,
                 Frith, MacMillan,
                 Owens, & Johnstone
                 (1994)

   12      552   Hauser & Scharfetter   * Ego psychopathological
                 (1990)                   scores

   13       34   Hoffmann & Kupper      * Negative symptoms
                 (1997)

   14       89   Jacobs, Wissusik,      * Psychiatric diagnosis
                 Collier, Stackman, &   * Work history
                 Burkeman (1992)        * Availability of
                                          disability allowance
                                        * Job interview skills

   15      342   Johnstone, Firth,      * Previous occupation
                 Lang, & Owens (1995)   * Family psychiatric
                                          history
                                        * Psychopathology

   16      237   Johnstone,             * Relapse rate
                 Macmillan, Frith,      * Pre-treatment duration
                 Lang, & Owens (1995)     of illness

   17       57   Lustman, Velozo,       * Psychiatric diagnosis
                 Eubanks, Montag, &
                 Cole (1991)

   18       91   Lysaker, Bell, Sito,   * Social skills
                 & Bioty (1995)         * Cognitive functioning

   19       68   Lysaker & Bell         * Social skills
                 (1995)

   20       60   Lysaker, Bell, &       * Cognitive functioning
                 Bioty (1995)

   21       35   Lysaker, Bell,         * Quality of social
                 Milstein Bryson,         contacts
                 Shestopal, &           * Previous social
                 Goulet (1993)            functioning

   22      402   Marneros, Deister &    * Diagnosis
                 Rohde (1992)

   23      143   Massel, Liberman,      * Psychiatric symptoms
                 Mintz, Jacobs, Rush,
                 Giannini, & Zarate
                 (1990)

   24      543   Mowbray, Bybee,        * Previous hospitalization
                 Harris, & McCrohan     * Attitude towards work
                 (1995)                 * Family relationship
                                        * Individual functioning
                                        * Service setting

   25       53   Munk-Jorgensen &       * Urban vs rural area
                 Mortensen (1992)

   26      309   Rabinowitz, Modai, &   * Premorbid functioning
                 Inbar-Saban (1994)     * Treatment in the hospital
                                        * Length of hospitalization
                                        * Neuropathology

   27      182   Razzano & Cook
                 (1994)

   28      200   Rimmerman, Botuck,     * Psychiatric diagnosis
                 & Levy (1995)          * Age
                                        * Gender

   29      275   Rogers, Anthony,       * Race
                 Cohen, & Davies        * Living arrangement
                 (1997)                 * Marital status
                                        * Occupational history
                                        * Psychiatric
                                          symptomatology
                                        * Substance abuse

   30      275   Rogers, Anthony,       * Marital status
                 Toole, & Brown         * Psychiatric
                 (1991)                   symptomatology
                                        * Work adjustment skills

   31       46   Solinski, Jackson,     * Negative symptomatology
                 & Bell (1992)          * Prior functioning
                                        * Interview skills

   32       87   Wohrl (1990)           * Baseline performance

   33      140   Van OS, Fahy, Jones,   * Neuropathology
                 Harvey, Lweis,         * Cognitive functioning
                 Williams, Toone &
                 Murray (1995)

   34      214   Vetter & Koller        * Severity of illness
                 (1996)                 * Social functioning
                                        * Age at first
                                          hospitalization

   35      258   Vogel, Blumenthal,     * Prior occupational
                 Neumann, Schuttler       history
                 (1988)

 Number                               Non-significant
of study        Author(s)               predictors

    1      Anthony, Rogers,       * Diagnosis
           Cohen, & Davies
           (1995)

    2      Bailer, J., Brauer,
           W., Rey, E.R. (1996)

    3      Beiser, Bean,
           Erickson, Zhang,
           Iacono, & Rector
           (1994)

    4      Bell & Lysaker         * Positive symptoms
           (1995)

    5      Breier, Schreiber,     * Positive symptoms
           Dyer & Pickar (1992)   * gender

    6      Carpenter & Strauss    * Cross-sectional
           (1991)                   symptoms

    7      Charisiou, Jackson,
           Byole, Burgess,
           Minas, & Joshua
           (1989)

    8      Corrigan, Reedy,       * Age
           Thadani, & Ganet       * Sex
           (1995)                 * Ethnicity
                                  * Marital status
                                  * Past work history
                                  * Education
                                  * Previous work history

    9      Gaebel & Pietzcker
           (1987)

   10      Gaebel & Pietzcker
           (1985)

   11      Geddes, Mercer,        * Psychopathology
           Frith, MacMillan,
           Owens, & Johnstone
           (1994)

   12      Hauser & Scharfetter   * Positive symptoms
           (1990)

   13      Hoffmann & Kupper      * Work performance at
           (1997)                   intake

   14      Jacobs, Wissusik,
           Collier, Stackman, &
           Burkeman (1992)

   15      Johnstone, Firth,
           Lang, & Owens (1995)

   16      Johnstone,
           Macmillan, Frith,
           Lang, & Owens (1995)

   17      Lustman, Velozo,
           Eubanks, Montag, &
           Cole (1991)

   18      Lysaker, Bell, Sito,
           & Bioty (1995)

   19      Lysaker & Bell
           (1995)

   20      Lysaker, Bell, &
           Bioty (1995)

   21      Lysaker, Bell,
           Milstein Bryson,
           Shestopal, &
           Goulet (1993)

   22      Marneros, Deister &
           Rohde (1992)

   23      Massel, Liberman,      * Psychiatric diagnosis
           Mintz, Jacobs, Rush,
           Giannini, & Zarate
           (1990)

   24      Mowbray, Bybee,
           Harris, & McCrohan
           (1995)

   25      Munk-Jorgensen &
           Mortensen (1992)

   26      Rabinowitz, Modai, &
           Inbar-Saban (1994)

   27      Razzano & Cook         * Sex
           (1994)

   28      Rimmerman, Botuck,
           & Levy (1995)

   29      Rogers, Anthony,       * Gender
           Cohen, & Davies        * Educational level
           (1997)                 * Previous
                                    hospitalizations
                                  * Criminal record
                                    diagnosis

   30      Rogers, Anthony,       * Race
           Toole, & Brown         * Gender
           (1991)                 * Educational status
                                  * Age
                                  * Residential status
                                  * Previous occupational
                                    history
                                  * Previous
                                    hospitalization
                                  * Diagnosis substance
                                    abuse

   31      Solinski, Jackson,     * Positive symptoms
           & Bell (1992)

   32      Wohrl (1990)           * Psychiatric diagnosis

   33      Van OS, Fahy, Jones,
           Harvey, Lweis,
           Williams, Toone &
           Murray (1995)

   34      Vetter & Koller
           (1996)

   35      Vogel, Blumenthal,
           Neumann, Schuttler
           (1988)

Table 2 Significant and non-significant predictors as identified from the studies reviewed

                                 Significant     Non-significant
                                 predictors as   predictors as
                                 indicated by     indicated by
                                  the number       the number
                                  of studies       of studies
Predictors                         reviewed         reviewed

                                       N                N

Medical aspects
  Diagnosis                            3                5
  Psychiatric symptomatology          10                6
  Neuropathology                       2                0
  Severity of illness                  1                0
  Use of neuroleptic drugs             1                0
  Age at first hospitalization         1                0
  Biological vulnerability             1                0
  Family psychiatric history           1                0
  Pre-treatment duration of
    Illness                            1                0

Demographic characteristics
  Age                                  1                2
  Sex                                  2                5
  Education                                             3
  Ethnicity                            1                2
  Marital status                       2                1
  Social class                         1                0
  Residential area                     2                1

Psychological aspects
  Cognitive functioning                4                2
  Ego strength/functioning             1                0

Social aspect
  Personal presentation                1                0
  Social skills                       11                0
  Social contact                       2                0
  Heterosexual relationship            1                0
  Family relationship                  1                0

Functional aspects
  Premorbid functioning (1)            6                0
  Premorbid occupational (2)
    performance                        7                3
  Attitude towards work                1                0
  Work adjustment skills               1                0

Environmental aspects
  Availability of employment
    assistance                         1                0
  Disability Allowance                 1                0
  Service in the hospital              1                0

Others
  Length of Hospitalization            2                0
  Substance abuse                      1                1
  Criminal record                      0                1
  Relapse rate                         1                0
  Previous hospitalization             1                2
  QOL                                  1                0

Footnotes (1) Premorbid functioning refers to functional level in work, leisure, and social aspects before the onset of mental illness. (2) Premorbid occupational performance refers to performance in job acquisition, work skills, work adjustment, and other related life experiences.

Discussion

Premorbid functioning

In this literature review, it is found that premorbid functioning, especially premorbid occupational performance, is a significant and consistent predictor of employment outcome of the psychiatric population (e.g. Breier, Schreiber, Dyer, & Pickar, 1994; Jacobs, Wissusik, Collier, Stackman, Burkeman, 1992; Rabinowitz, Modai, & Inbar-Saban, 1994). For example, Beiser et al. (1994) studied the occupational functioning of a group of patients with schizophrenia (N=31) or affective psychosis (N=33) 18 months after the first episode using a variety of sociodemographic, clinical, and psychophysiological measures. They found that premorbid job performance was one of three good occupational outcome predictors. Also, researchers (Rabinowitz et al., 1994) attempted to identify variables associated with improved functioning of 309 patients in an open ward over two years after psychiatric hospitalization using stepwise logistic regression. They found that preadmission level of functioning was found to be one of the best predictors. Similarly, Jacobs et al. (1992) studied a group of 89 in-patient and community-dwelling psychiatric patients in the Los Angeles area who participated in a job-finding club. They found that those with good work histories were more likely to find employment. These results are in line with many previous studies (Anthony & Jansen, 1984; Stephens, Astrup, & Mangrum, 1966; Strauss & Carpenter, 1974; Vaillant, 1964). Although premorbid functioning is shown to be a consistent predictor, studies indicated that it does not account for all variance in outcome. For example, Breier et al. (1992) found that intake occupational functioning, plexus visibility, and premorbid social functioning accounted only for 43% of the variance in 18-month outcome for the study's schizophrenic subjects. As a result, other strong predictors that account for perhaps larger amounts of variance should also be identified.

According to Anthony and Jansen (1984), the best demographic predictor of vocational outcome among persons with psychiatric illnesses is previous employment history. For instance, an employed group of people with schizophrenia was shown to have better improvement in work skills than those who had never been employed (Anthony, Rogers, Cohen, & Davies, 1995). The unemployed group was shown to be lacking in relevant life experiences, which in turn led to reduced chances to learn the demands of the working world and to reduced work adjustment skills (Ruttman, 1994). Based on these consistent results, therapists and vocational counselors should not overlook this aspect when planning vocational rehabilitation for those suffering from mental illness. Although the present literature review shows that there are some studies which find that previous work history is not a significant correlate of work performance, these results are not strong enough to rule out the general pattern as illustrated above. These apparently contradictory results may be due to methodological reasons: sampling bias and special operational definitions of these studies. For example, in a study by Rogers, Anthony, Toole, and Brown (1991), it was found that the previous work histories of a group of 275 participants in psychosocial rehabilitation did not correlate significantly with their employment outcomes. On closer examination, however, it is found that previous work history was the main indicator of premorbid functioning. In their study, work history was defined in terms of the number of months employed during the previous five years. This measure may not be valid and representative of the participants' premorbid functioning. This result is contrary to the majority of other studies. Thus, this problem with the measure may explain the contrary results.

Social skills

The significance of social skills for people with chronic mental illness in functioning effectively in the community, including the workplace, has been extensively discussed (Tsang & Pearson, 1996). In brief, this model conceptualizes work-related social skills as a three-tier structure in which the three tiers have a hierarchical relationship. From a learning point of view, a person should first master concepts and skills in the fundamental stage before proceeding to the next. The first tier comprises basic social skills and basic social survival skills. Basic social skills focus on interpersonal communication. These skills relate to the receiving, processing, and sending of information as defined by Liberman, Mueser, Wallace, Jacobs, and Mueser (1986). Basic social survival skills consist of such skills as grooming, politeness, and personal appearance. The second tier comprises two clusters of core skills. Core skills are defined as those needed in handling both general and specific work-related situations. The first cluster includes those skills required for coping with any job irrespective of its specific nature. General work-related skills can be divided into job-securing social skills and job-retaining social skills. The latter set of skills may further be divided according to the three groups of people with whom an employee must interact: supervisors, colleagues, and subordinates. The second cluster of core skills consists of skills vital for coping with situations specific to a particular kind of job. For instance, a receptionist must possess skills necessary for dealing with inquiries from visitors or customers of a company; a sales person in a clothing store has to know how to sell the store's products and how to cater to the needs of the customers. Finally, the third tier of the model encompasses the goals to which the basic and core skills point: in other words, the benefits that a person can obtain by possessing these skills. These benefits embrace getting a job, settling into a job, maintaining a job, and deriving a sense of achievement and satisfaction from the job.

Social skills are essential for individuals to develop supportive social networks, cope with day to day stress, improve social functioning, and find and hold a job. The results of this study support previous findings that social skill or social functioning level is a strong and consistent predictor of vocational outcome. Support for this position was obtained from 11 research projects with different methodologies as reviewed in this study (e.g., Bell & Lysaker, 1995; Carpenter & Strauss, 1991; Charisiou et al., 1989; Lysaker, Bell, & Bioty, 1995; Lysaker & Bell, 1995; Rutman, 1994; and Solinski, Jackson, & Be11,1992). Charisiou et al. (1989) examined the effects of verbal and nonverbal interview microbehaviours and interview characteristics on the employability of a group of 46 in-patients with schizophrenia. They discovered that manifest adjustment and ability to communicate accounted for 64% of the total variance in predicting employability. Solinski et al. (1992) conducted a study at a major metropolitan state psychiatric hospital in Victoria, Australia using the more advanced statistical technique of LISREL. It was found that social skills show very high correlation with overall employability. In another longitudinal 14-year follow-up study conducted in Germany (Vetter & Koller, 1996), it was found that social functioning was one of the three strongest predictors of occupational development in a group of 214 psychiatric patients. In fact, studies conducted recently in Hong Kong also revealed similar findings. A survey carried out among medical and rehabilitation professionals (N=118) at a mental hospital showed that it was generally perceived that social competence in general and in the workplace were strong predictors of vocational outcome (Tsang, Lam, Darasi, & Ng, in press). In another study by Tsang (1996), a survey was conducted among 44 rehabilitation professionals and 54 clients with schizophrenia in halfway houses and sheltered workshops. The majority of the respondents thought that being socially competent was an important factor for successful employment for people suffering from schizophrenia.

Psychiatric symptomatology and diagnosis

One of the interesting results in the present study is that psychiatric symptomatology (which refers to the abnormalities in moods, thoughts and behaviors resulting from mental illness) and diagnosis are found to be significant predictors of employment outcome. This result is in contradiction of some of the earlier studies as summarized by Anthony and Jansen (1984).

In fact, opposing views on the relationship between psychiatric symptoms and work function have existed for quite a long time and are well known among researchers. Bell and Lysaker (1995) commented that this difference has great implications on management of clients and needs to be addressed. Before the mid 1980s, the limited research information in this area tended to support the view that symptoms and work are unrelated (Anthony & Jansen, 1984). It seems that this belief is challenged by the current review. Altogether 10 out of 35 studies indicate that there existed a relationship between these two variables. Studies supporting psychiatric symptomatology include Anthony et al. (1995), Bell & Lysaker (1995), and Solinski et al. (1992). Anthony et al. (1995) found that symptomatology was a highly significant predictor, especially when using the more stringent definition of employment as having had full-time employment for 12 weeks. A similar finding was obtained by Bell and Lysaker (1995) in which psychiatric symptoms were shown to be useful predictors for work performance. The two sets of studies may have different findings because the early studies had small numbers of clients with psychiatric disabilities and the clients' motivations for work were not well considered.

Similar to research by Mowbray et al. (1995), another reason for the conflicting results may be due to the vague definitions given to psychiatric symptoms. In fact, the failure to use consistent dimensions in defining symptoms is a common feature among previous outcome studies (Breier et al., 1992). When psychiatric symptoms are further classified into either positive or negative, however, the results are no longer as contradictory as they first appear. According to the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (American Psychiatric Association, 1994), positive symptoms reflect an excess or distortion of normal functions (e.g., hallucination, delusion, etc.) while negative symptoms reflect a diminution or loss of normal functions (e.g., social withdrawal, avolition, etc.). A closer look shows that those studies in support of symptomatology as a significant predictor of vocational outcome referred to negative symptoms (e.g., Bailer, Brauer, & Rey, 1996; Bell & Lysaker, 1995; Hoffman & Kupper, 1997). On the other hand, those studies not in support of symptomatology as a significant predictor referred to positive symptoms (e.g., Bell & Lysaker, 1995; Hauser & Scharfetter, 1990; Solinski et al., 1992).

In a study by Solinski et al. (1992) on a group of 46 chronic in-patients with schizophrenia, it was found that negative symptoms exerted a substantial influence on the sample's employability ratings. However, positive symptoms, including delusions, thought disorders and bizarre behaviors, exerted little effect on employment. In the study of Anthony et al. (1995), it was found that there was a stronger and more consistent relationship between negative symptoms and work skills (a range of -.13 to -.39) than between positive symptoms and work skills (a range of -.05 to -.24). In this study on 51 first-onset schizophrenia patients, it was found that poor occupational outcome was not significantly associated with any psychopathology predictors at five-year follow up. One explanation of this result may be that the first episode of schizophrenia will usually not have marked negative symptoms. Bell and Lysaker (1995) conducted a study of 61 subjects with a DSMIII diagnosis of schizophrenia and schizoaffective disorder. This study used the positive- and negative-symptoms scale to predict performance on the work personality profile at 3 to 13 weeks after attending a psychosocial rehabilitation program. It was found that 30% of work performance was explained by components including cognitive function, hostility, and negative symptoms. Again positive symptom components did not predict work performance. Breier et al. (1992) studied the outcome of a group 58 patients suffering from chronic schizophrenia admitted to the NIH Clinical Center and found that positive symptoms as measured by the Brief Psychiatric Rating Scale (Overall & Gorham, 1961) did not predict outcome. To conclude, based on the above analysis, it can be postulated that the conflicting results between the more recent studies and earlier studies may be resolved if a clearer definition of psychiatric symptoms, particularly positive vs. negative, is used. It is negative symptoms, not positive, that affect work functioning and job interview performance, which in turn predicts employment outcome. Further research may be planned to examine the differential effect of positive and negative symptoms on employability. Additional research may be the best way to resolve this controversial issue on the relationship between psychiatric symptomatology and vocational outcome.

Based on our results and previous studies, it is still not certain whether or not psychiatric diagnosis is a good and significant predictor of employment outcome for people with psychiatric disabilities. According to some early studies (e.g., Distefano & Pryer, 1970; Goss & Pate, 1967) and the classic review conducted by Anthony and Jansen (1984), diagnosis is a poor predictor. Five of the studies reviewed came to the same conclusion. Out of these five, three belonged to the same series, based on the same sample of research participants (Anthony et al., 1995; Rogers et al., 1991, 1997). These three articles described related studies to examine the clinical and demographic correlates of work skills and vocational outcome for 275 vocationally ready persons with psychiatric disabilities. Of these 275 clients, 157 (57.1%) suffered from schizophrenia; 47 (17.1%) suffered from bipolar disorder or major depression; 54 (19.6%) suffered from anxiety, personality, paranoid, dysthymic, or developmental disorder; and the diagnoses of the remaining 17 (6.2%) clients were unknown. The findings were that diagnostic category was not predictive of work outcome. Similarly, Massel and his colleagues (1990) explored the relationship between psychiatric symptomatology and the capacity to work in 143 mentally ill clients. They found that there was no significant difference in the work performance between the non-psychotic and the psychotic clients.

On the other hand, Jacobs and his colleagues (1992) studied the correlation between psychiatric disabilities and vocational outcome of 89 clients attending a job-finding club program. They found that persons with non-psychotic diagnoses were more likely to find employment. Also, there was a study to predict the likelihood of job placement in a supported employment program for 200 subjects with either psychiatric disabilities, severe learning disabilities, or mental retardation (Rimmerman, Botuck, & Levy, 1995). It was found that, overall, individuals with psychiatric abilities had a lower chance to be placed than patients with either learning disabilities or mental retardation.

Based on the above findings, it seems that the latest findings do not shed much light on the existing argument about the ability of psychiatric diagnosis to predict employment outcome. The above studies reviewed shared some common limitations. First, psychiatric diagnosis was not a main predictor variable to be examined. In most of the studies, diagnosis was only a side variable to be investigated along with other variables such as psychiatric symptomatology. Second, the diagnostic criteria and process had generally not been consistently defined or clearly spelled out. Third, the range of diagnoses to be compared was not broad enough to represent the whole spectrum of mental illnesses Finally, the sample size of different diagnoses was significantly unequal, with schizophrenia and psychoses usually outnumbering other types of diagnosis. As a result, research that primarily aims at studying the predictive value of psychiatric diagnosis on employment outcome of people with mental illness is urgently needed. In this research, sample size should be large enough to equally represent patients with different diagnoses. Diagnostic criteria should be clearly defined in accordance with international practice. Also, procedures should be established so that the process of making the diagnosis is reliable and valid.

Cognitive functioning

Cognitive rehabilitation emerged as a new approach in treating persons with severe mental illness more than a decade ago (Penn, 1991). The approach focuses on the client's competencies and weaknesses and applies intervention strategies often directly derived from the measures used for assessment. Cognitive deficits of schizophrenia, especially impairment in attention span, have been extensively documented (Elliott & Sahakian, 1995; Tryssenaar & Goldberg, 1994). However, the relationship between cognitive function and work performance among individuals with mental illness has been relatively neglected in earlier research on employment outcome.

In this study, cognitive function was found to be a significant indicator of employment outcome in four of the studies reviewed. Three of them belong to the same series based on the same group of research participants, conducted by Lysaker and his colleagues (Bell & Lysaker, 1995; Lysaker, Bell, Bioty, 1995; Lysaker et al., 1995). These three studies originated from the observation that, for a subgroup of patients, their symptoms remained stable or even worsened while participating in work rehabilitation program. This subgroup of patients did not respond to the treatment. Lysaker and his colleagues hypothesized that this group of patients shared a common characteristic: cognitive impairment. In order to test their hypothesis, they recruited subjects with schizophrenia or schizoaffective disorder, as defined in the Diagnostic and Statistical Manual of Mental Disorders (Third Edition) Revised, for a work rehabilitation study. After the recruitment process, the researchers used (a) the Positive and Negative Syndrome Scale (PANSS; Kay et al., 1987); (b) the Wisconsin Card Sorting Test (Heaton, Chelune, Talley, Kay, & Curtiss, 1993), which is a neuro-psychological test sensitive to difficulties in information processing, concept formation, and flexibility of abstract thought; and (c) the Slosson Intelligence Test (Slosson, 1962); to measure cognitive impairment. Using multiple regression analysis to determine whether work performance of the participants could be predicted by PANSS component scores, these researchers found that the cognitive and negative components most frequently accounted for the variance of work performance. These components contributed to 14 of 16 significant predictor equations (Bell & Lysaker, 1995). Discriminant analysis revealed that cognitive impairments predicted improvement in the participants in the work rehabilitation program (Lysaker, Bell, & Bioty, 1995). Multiple regression analyses indicated cognitive impairment was associated with social skills at work, which would then predict the work performance of the subjects (Lysaker, Bell, Sito & Bioty, 1995). The fourth is a study to examine the relationship between intra- and extracerebral cerebrospinal fluid and unemployment of 140 patients suffering from psychotic illness (Van OS et al., 1995). In this study, cognitive functioning was found to be a mediating factor along the causal pathway between ventricular enlargement and unemployment. This result is supported by the finding that very significant associations exist between sylvian fissure volumes and a measure of cognitive functioning at follow-up and that controlling for cognitive functioning greatly reduced the strength of the association with unemployment. According to the authors, this finding is consistent with previous studies (e.g., Golden, Moses, & Zelazowski, 1980).

It was suggested that neuro-psychological impairment was associated with poorer work performance. Although with methodological flaws, research establishes the pattern that cognitive deficit is highly likely to be related to an impairment in vocational functioning. More research effort should, therefore, be geared towards clarification of this issue. Recently, the focus has been in the social cognition of schizophrenia and other mental illness (Penn, Corrigan, Bentall, Racenstein, 1997). Social cognition refers to the domain of cognition that involves the perception, interpretation, and processing of social information (Ostrom, 1984). The deficits in social cognition are believed to contribute to poor premorbid social competence. A growing body of evidence shows that impairments in affect perception and social cognitive problem solving are common in people with schizophrenia and are associated with their deficits in social competence (Bellack, Morrison, Mueser, 1989; Mueser et al., 1996). As previously mentioned, studies have indicated that social competence is a strong predictor of employment outcome. To follow this line of thought, although lacking in empirical evidence, impairment in social cognition should also have a close relationship with employment outcome of patients with schizophrenia and psychoses. Researchers, therefore, should direct their efforts to study this hypothesis with appropriate controls.

Family relationship

Family relationship has been well accepted as an important element of psychiatric rehabilitation. Though there has been an expansion of research studies on family burdens and family intervention programs, this study found that scant attention has been given to the effect of family functioning on the employment outcome of persons with mental illness. In this review, only one of 35 studies was found to have assessed the relationship between family relationship and employment outcome. The study by Mowbray et al. (1995) showed that, in a sample of 437 individuals with a severe mental illness selected in two case management agencies in Michigan, vocational outcome was related to the frequency of family contact as measured by the Community Living Adaptation Scale (CLAS) ratings. In their study, it was found that those who reported doing things with their families once a month or less were more than twice as likely to be employed as those who reported more frequent family contact. According to the authors, the causal direction of this relationship and the underlying reason are unknown. Two possible reasons may explain this special phenomenon. First, more contact outside the family may reflect sociability or independent functioning and thus a greater likelihood of finding employment. Second, employment may lead to less dependency on the family to meet social or instrumental needs. Further research needs to clarify this interesting finding.

In fact, the relationship between family and relapse has been a topic of intensive study for the past two decades, following the pioneering work on expressed emotion (Brown, 1985). Many interventions programs are based on the research work on expressed emotions. They are found to be effective in reducing the relapse rate of persons with schizophrenia. Families are seen as an essential part of support and care for discharged patients with mental illness. Logically, quality and quantity of family contacts make a difference in the vocational outcome of persons with mental illness. When such a relationship is made clear, interventions may be developed to increase the employability of this group of persons. The connection between family relationship and vocational outcome among persons with mental illness is a neglected area that needs to be addressed in future psychiatric rehabilitation research.

Substance Abuse

The literature shows that the importance of employment and employment outcome is well recognized in the field for rehabilitation for drug abusers (Platt, 1995). By securing and holding a job, drug abusers can establish a legal source of income; may improve their self-esteem, reducing the likelihood of using illegal drugs; and may refrain from criminal activity (Faupel, 1988; Joe, Chastain, & Simpson, 1990). Platt points out that employment has been used as a criterion for treatment outcome in addiction treatment evaluation research. In 31 studies reviewed by McLellan (1983), 26 were found to have included employment status as a significant correlate of retention in treatment.

For people with schizophrenia, research shows that their incidence rate of substance abuse has been as high as 29 to 50% (Kirchner, Owen, Nordquist, & Fischer, 1998). Research (e.g., Barbee, Clark, & Crapanzano, 1989) has shown that the existence of substance abuse in schizophrenia may contribute to poor treatment outcome. However, research to examine the effect of substance abuse on employment outcomes of people with mental illness has been limited. Although studies to identify demographic and clinical correlates of employment outcome have been voluminous, the inclusion of substance abuse as a variable among these studies is rare. In this review, only two studies of the 35 included drug abuse as a dependent variable. Rogers, Anthony, Cohen, & Davies, (1997) pointed out that alcohol or substance abuse problems accounted for 1% of work performance of 275 clients with psychiatric disabilities. In another paper based on the same sample (Rogers et al., 1991), the authors compared the employed and unemployed subjects among those who had been in the study for one year or more. They found that, although not statistically significant, alcohol or substance abuse problems nearly approached significance (chi-square=3.55, p=.06). In both articles, the main focus was to resolve whether psychiatric symptomatology and diagnosis were significant predictors. These studies used a longitudinal approach as a relatively new research methodology rather than taking substance abuse as a main dependent variable to see whether it is a significant predictor of vocational outcome of people with psychiatric disabilities. It is therefore recommended that controlled studies be designed and implemented to closely examine the relationship between drug abuse and employment outcomes in persons suffering from severe mental illness.

Implications

To summarize, this research literature review looked at 35 controlled studies on identification of good and poor predictors of vocational outcome for patients with psychiatric disabilities conducted between 1985 and 1997. In line with previous findings, the results show that premorbid functioning, work history, and social skills are good and significant predictors. Symptomatology and diagnosis, however, continue to show contradictory results. Suggestions are made to resolve the contradiction. Also, some relatively neglected aspects, such as cognitive functioning and family relationship, are found to be significant predictors. These areas deserve more attention in future research. It is suggested that more controlled and empirical studies should be conducted in Hong Kong to verify whether or not predictors identified and reviewed in this literature research are really significant in predicting employment outcome of discharged psychiatric patients.

This review does not merely shed light on future research directions in the area of vocational rehabilitation. Instead, the review has significant implications on the design and formulation of vocational rehabilitation for people with psychiatric disabilities. Models for vocational rehabilitation for people with disabilities have recently been reviewed (e.g., Chan, et al., 1997; Bolton, 1988; Innes & Straker, 1998). Although models emphasize different areas, common areas of assessment include physical aspects, such as muscle strength and balance, and functional aspects, such as lifting and manual dexterity, personality, interest, and job interview skills. Using the results of this study, therapists and vocational counselors working in psychiatric rehabilitation settings should assess factors which are shown to be significant predictors of employment for people with mental illness. Examples of such factors are premorbid functioning, occupational history and social skills. In order to obtain reliable and valid data, standardized tests and checklists should be used. For example, the premorbid functioning could be assessed by the Life Skills Profile (LSP; Rosen, Hadzi-Pavlovic, & Parker, 1989) which is a 39-item scale of personal functioning completed by a clinician. Similarly, social skills necessary for seeking and keeping a job may be evaluated by a two-part measure developed and validated by Tsang and Pearson (b) (in press). With this information, the vocational potential of the clients can be more accurately predicted.

Knowledge of the employability and placeability of their mentally ill clients based on a comprehensive vocational evaluation helps professionals choose the best vocational rehabilitation strategies. There are a wide range of vocational rehabilitation programs for people with psychiatric disabilities: hospital-based programs, sheltered workshops, assertive case management, psychosocial rehabilitation, transitional employment, and supported employment (Lehman, 1995). For clients who show potential for successful entry into competitive employment, appropriate training strategies such as transitional employment (Malamud & McCrory, 1988), case management (Mueser et al., 1997) and social skills training (Liberman, Mueser, & Wallace, 1986; Mueser & Liberman, 1988; Tsang & Pearson (a) (in press) would be most appropriate. On the other hand, supported employment (Drake, 1998) may be arranged for those who have lower potential for competitive employment.

With these empirical data, it is hoped that therapists and counselors will be able to design and conduct more effective rehabilitation programs. In addition, we expect that more research effort may be directed in clarifying such intriguing findings as the connection between family relationship and employment outcome revealed in this study.

References

Anthony, W. A., Cohen, M. R., & Vitals, R. I. (1978). The measurement of rehabilitation outcome. Schizophrenia Bulletin, 4, 365-383.

Anthony, W. A., & Jansen, M. A. (1984). Predicting the vocational capacity of the chronically mentally ill: Research and policy implications. American Psychologist, 39(5). 537-544.

Anthony, W. A., Rogers, E. S., Cohen, M., & Davies, R. R. (1995). Relationships between psychiatric symptomatology, work skills, and future vocational performance. Psychiatric Services, 46(4), 353-358.

Bailer, J., Brauer, W., & Rey, E. R. (1996). Premorbid adjustment as predictor of outcome in schizophrenia: Results of a prospective study. Acta Psychiatr Scand, 93, 368-377.

Barbee, J. G., Clark, P. D., Crapanzano, M. S. (1989). Alcohol and substance abuse among schizophrenic patients presenting to an emergency psychiatric service. Journal of Nervous and Mental Disease, 177, 400-407.

Beiser, M., Bean, G., Erickson, D., Zhang, J., Iacono, W. G., & Rector, N. A. (1994). Biological and psychosocial predictors of job performance following a first episode of psychosis. American Journal of Psychiatry, 151(6), 857-863.

Bell, M. D., & Lysaker, P. H. (1995). Psychiatric symptoms and work performance among persons with severe mental illness. Psychiatric Services, 46(5), 508-510.

Bellack, A. S., Morrison, R. L., & Mueser, K. T. (1989). Social problem solving in schizophrenia. Schizophrenia Bulletin, 15, 101-106.

Bolton, B. (1988). Vocational assessment of persons with psychiatric disorders. In J. A. Ciardiello & M. D. Bell (Eds), Vocational rehabilitation of persons with prolonged psychiatric disorders (pp. 165 - 180). Baltimore: The Johns Hopkins University Press.

Bond, G. R. (1992). Vocational rehabilitation. In R. P. Liberman (ed.). Handbook of psychiatric rehabilitation (Vol. 166) (pp. 244-275). Boston: Allyn & Bacon.

Breier, A., Schreiber, J. L., Dyer, J., & Pickar, D. (1992). Course of illness and predictors of outcome in chronic schizophrenia: Implications for pathophysiology. British Journal of Psychiatry, 161(Suppl. 18), 38-43.

Brown, G. W. (1985). The discovery of expressed emotion: induction or deduction. In J. Leff & C. Vaughn (eds.). Expressed emotion in families. New York: Guilford.

Carpenter, W. T., & Strauss, J. S. (1991). The prediction of outcome in schizophrenia IV: Eleven-year follow-up of the Washington IPSS cohort. The Journal of Nervous and Mental Disease, 179(9), 517-525.

Chan, F., Reid, C., Kaskkel, L. M., Roldan, G., Rahimi, M., Pmofu, E. (1997). Vocational assessment and evaluation of people with disabilities. Physical Medicine and Rehabilitation Clinics of North America, 8(2), 311-325.

Charisiou, J., Jackson, H. J., Boyle, G. J., Burgess, P. M., Minas, I. A., & Joshua, S. D. (1989). Which employment interview skills best predict the employment of schizophrenic patients? Psychological Reports, 64, 683-694.

Corrigan, P. W., Reedy, P., Thadani, D., & Ganet, M. (1995). Correlates of participation and completion in a job club for clients with psychiatric disability. Rehabilitation Counseling Bulletin, 39(1), 42-53.

Dawis, R. (1987) A theory of work adjustment. In B Bolton (ed). Handbook on the measurement and evaluation in rehabilitation, 2 ed. (pp. 207-217). Baltimore, MD: Paul H. Brooks.

Dion, G. L., & Anthony, W. A. (1987). Research in psychiatric rehabilitation: A review of experimental and quasi-experimental studies. Rehabilitation Counseling Bulletin, March, 177-203.

Distefano, M. K., & Pryer, M. W. (1970). Vocational evaluation and successful placement of psychiatric clients in a vocational rehabilitation program. American Journal of Occupational Therapy, 24, 205-207.

Drake, R. E. (1998). A brief history of the individual placement and support model. Psychiatric Rehabilitation Journal, 22 (1), 3-7.

Elliott, R., & Sahakian, B. J. (1995). The neuropsychology of schizophrenia: relations with clinical and neurobiological dimensions. Psychological Medicine, 25, 581-594.

Faupel, C. E. (1988). Heroin use, crime and employment status. The Journal of Drug Issues, 18(3), 467-479.

Gaebal, W., & Pietzcker, A. (1985). Multidimensional study of the outcome of schizophrenic patients one year after clinical discharge: Predictors and influence of neuroleptic treatment. Eur Arch Psychiatr Neurol Sci, 235, 45-52.

Gaebal, W., & Pietzcker, A. (1987). Prospective study of course of illness in schizophrenia: Part III Treatment and outcome. Schizophrenia Bulletin, 13(2), 307-316.

Geddes, J., Mercer, G., Frith, C. D., Macmillan, F., Owens, D. G. C., Johnstone, E. C. (1994). British Journal of Psychiatry, 165, 664-668.

Golden, C. J., Moses, J., Zelazowski, R. (1980). Cerebral ventricular size and neuropsychological impairment in young schizophrenics. Archives of General Psychiatry, 37, 619.

Goldstrom, I., & Manderscheid, R. (1982). The chronically mentally ill: A descriptive analysis from the uniform client data instrument. Community Support Services Journal, 2, 4-9.

Goss, A. M., & Pate, K. D. (1967). Predicting vocational rehabilitation success for psychiatric patients with psychological tests. Psychological Reports, 21, 725-730.

Griffiths, R. D. P. (1977). The prediction of psychiatric patients' work adjustment in the community. British Journal of Social and Clinical Psychology, 16, 165-173.

Hall, J. C., Smith, K., & Shimkunas, A. (1966). Employment problems of schizophrenic patients. American Journal of Psychiatry, 123(5), 536-540.

Hauser, R., & Scharfetter, C. (1990). On the prognostic relevance of ego-psychopathology in schizophrenia: A 2.5 year follow-up. Eur Arch Psychiatr Neurol Sci, 239, 293-302.

Heaton, R. K., Chelune, G. J., Talley, J. L., Kay, G. C., & Curtiss, G. (1993). Wisconsin card sorting test manual revised and expanded. USA: Psychological Assessment Resources, Inc.

Hoffmann, H., & Kupper, Z. (1997). Relationships between social competence, psychopathology and work performance and their predictive value for vocational rehabilitation of schizophrenic outpatients. Schizophrenia Research, 23, 69-79.

Innes, E., & Straker, L. (1998). A clinician's guide to work-related assessments: 3 - Administration and interpretation problems. Work, 11, 207-219.

Jacobs, H. E., Wissusik, D., Collier, R., Stackman, D., & Burkeman, D. (1992). Correlations between psychiatric disabilities and vocational outcome. Hospital and Community Psychiatry, 43(4), 365-369.

Joe, G. W., Chastain, R. L., & Simpson, D. W. (1990). Relapse. In D. D. Simpson & S. B. Sells (Eds), Opioid addiction and treatment: A 12-year follow-up (pp. 121-136). Malabar, FL: Krieger.

Johnstone, E. C., Frith, C. D., F. H. Lang, & Owens, D. G. C. (1995). Determinants of the extremes of outcome in schizophrenia. British Journal of Psychiatry, 167, 604-609.

Johnstone, E. C., Macmillan, J. F., Frith, C. D., Benn, D. K., & Crow, T. J. (1990). Further investigation of the predictors of outcome following first schizophrenic episodes. British Journal of Psychiatry, 157, 182-189.

Kay, S. R., Fiszbein, A., & Opler, L. (1987). The positive and negative syndrome scale for schizophrenia. Schizophrenia Bulletin, 13, 261-276.

Kirchner, J. E., Owen, R. R., Nordquist, C., & Fischer, E. P. (1998). Diagnosis and management of substance use disorders among inpatients with schizophrenia. Psychiartric Services, 49 (1), 82-85.

Lehman, A. F. (1995). Vocational rehabilitation in schizophrenia. Schizophrenia Bulletin, 21 (4), 645-656.

Liberman, R. P., Mueser, K. T, & Wallace, C. J. (1986). Social skills training for schizophrenics at risk for relapse. American Journal of Psychiatry, 143, 523-526.

Lustman, P. J., Velozo, C. A., Eubanks. B., Montag, J. A., & Cole, D. M. (1991) Psychiatric disorder: Effects on rehabilitation and ability to return to work. Work, 37-43.

Lysaker, P., & Bell, M. (1995). Negative symptoms and vocational impairment in schizophernia: Repeated measurements of work performance over six months. Acta Psychiatr Scand, 91, 205-208.

Lysaker, P., Bell, M., & Bioty, S. M. (1995). Cognitive deficits in schizophrenia: Prediction of symptom change for participators in work rehabilitation. The Journal of Nervous and Mental Disease, 183(5), 332-336.

Lysaker, P., Bell, M., Milstein, R., Bryson, G., Shestopal, A., & Goulet, J. B. (1993). Work capacity in schizophrenia. Hospital and Community Psychiatry, 44(3), 278-280.

Lysaker, P., Bell, M., Sito, W. S., & Bioty, S. M. (1995). Social skills at work: Deficits and predictors of improvement in schizophrenia. The Journal of Nervous and Mental Disease, 183(11), 688-692.

Malamud, T. J., & McCrory, D. J. (1988). Transitional employment and psychosocial rehabilitation. In J. A. Ciardiello & M. D. Bell (Eds), Vocational rehabilitation of persons with prolonged psychiatric disorders (pp. 150 - 162). Baltimore: The Johns Hopkins University Press.

Marneros, A., Deister, A., & Rohde, A. (1992). Comparison of long-term outcome of schizophrenic, affective and schizoaffective disorders. British Journal of Psychiatry, 161 (Suppl. 18), 44-51.

Massel, H. K., Liberman, R. P., Mintz, J., Jacobs, H. E., Rush, T. V., Giannini, C. A., Zarate, R. (1990). Evaluating the capacity to work of the mentally ill. Psychiatry, 53, 31-43.

McCue, M., & Katz, G. L. (1983). The severely disabled psychiatri patient and the adjustment to work. Journal of Rehabilitation, 49, 52-58.

McLellan, A. T. (1983). Patient characteristics associated with outcome. In J. R. Cooper, F. Altman, B. S. Brown, & D. Czechowicz (eds.), Research on the treatment of narcotic addiction: State of the art (pp. 500-529). Washington, DC: U.S. Government Printing Office.

Mowbray, C. T., Bybee, D., Harris, S. N., & McCrohan, N. (1995). Predictors of work satus and future work orientation in people with a psychiatric disability. Psychiatric Rehabilitation Journal, 19(2), 17-28.

Mueser, K. T., Doonan, B., Penn, D. L., Blanchard, J. J., Bellack, A. S., Nishith, P., DeLeon, J. (1996). Emotion perception and social competence in chronic schizophrenia. Journal of Abnormal Psychology, 105, 271-275.

Mueser, K. T., & Liberman, R. P. (1988). In J. A. Ciardiello & M. D. Bell (Eds), Vocational rehabilitation of persons with prolonged psychiatric disorders (pp. 81 - 103). Baltimore: The Johns Hopkins University Press.

Munk-Jorgensen, & Mortensen, S. B. (1992). Social outcome in chizophernia: A 13-year follow-up. Social Psychiatry and Psychiatric Epidemiology, 27, 129-134.

Osipow, S. H. (1968). Theories of career development. New York: Appleton-Century-Crofts.

Ostrom, T. M. (1984). The sovereignty of social cognition. In R. S. Wyer & T. K. Srull (Eds), Handbook of social cognition (Vol. 1, pp. 1-37). Hillsdale, NJ: Erlbaum.

Overall, J. E., & Gorham, D. E. (1961). The Brief Psychiatric Rating Scale. Psychological Reports, 10, 799-812.

Penn, D. L. (1991). Cognitive rehabilitation of social deficits in schizophrenia: A direction of promise or following a primrose path. Psychosocial Rehabilitation Journal, 15, 1, 27-41.

Penn, D. L., Corrigan, P. W., Bentall, R. P., & Racenstein (1997). Social cognition in schizophrenia. Psychological Bulletin, 121 (1), 114-132.

Platt, J. J. (1995). Vocational rehabilitation of drug abusers. Psychological Bulletin, 117, 3, 416-433.

Rabinowitz, J., Modai, I., & Inbar-Saban, N. (1994). Understanding who improves after psychiatric hospitalization. Acta Psychiatr Scand, 89, 152-158.

Razzano, L., & Cook, J. A. (1994). Gender and vocational assessment of people with mental illness: What works for men may not work for women. Journal of Applied Rehabilitation Counseling, 25(3), 22-31.

Rimmerman, A., Botuck, S., & Levy, J. M. (1995). Job placement for individuals with psychiatric disabilities in supported employment. Psychiatric Rehabilitation Journal, 19(2), 38-43.

Rogers, E. S., Anthony, W. A., Cohen, M., & Davies, R. R. (1997). Prediction of vocational outcome based on clinical and demographic indicators among vocationally ready clients. Community Mental Health Journal, 33(2), 99-112.

Rogers, E. S., Anthony, W. A., Toole, J., & Brown, M. A. (1991). Vocational outcomes following psychosocial rehabilitation: A longitudinal study of three programs. Journal of Vocational Rehabilitation, 1(3), 21-29.

Rosen, A., Hadzi-Pavlovic, D., & Parker, G. (1989). The Life Skills Profile: A measure assessing function and disability in schizophrenia. Schizophrenia Bulletin, 15(2), 325-337.

Rutman, I. D. (1994). How psychiatric disability expresses itself as a barrier to employment. Psychosocial Rehabilitation Journal, 17(3), 15-35.

Slosson, R. (1962). Slosson intelligence test for children and adults. East Aurona, New York: Slosson Educational Publications.

Solberg, N. A., & Chueh, W. (1976). Performance in occupational therapy as a predictor of successful prevocational training. The American Journal of Occupational Therapy, 30(8), 481-486.

Solinski, S., Jackson, H. J., & Bell, R. C. (1992). Prediction of employability in schizophrenic patients, 7, 141-148.

Stauffer, D. L. (1986). Predicting successful employment in the community for people with a history of chronic mental illness. Occupational Therapy in Mental Health, 31-49.

Stephens, J. H., Astrup, C., & Mangrum, J. C. (1966). Prognostic factors in recovered and deteriorated schizophrenics. American Journal of Psychiatry, 122, 1116-1121.

Tryssenaar, J., & Goldberg, J. O. (1994). Improving attention in a person with schizophrenia. Canadian Journal of Occuapational Therapy, 61, 4, 198-205.

Tsang, W. H. H. (1995). The current scene of vocational rehabilitation for people with mental illness in Hong Kong, Hong Kong Journal of Mental Health, 24, 25-39.

Tsang, W. H. H., Lam, P., Dasari, B., & Ng, B. (in press). Predictors of post-hospital employment status for psychiatric patients: A survey among medical and rehabilitation professionals.

Tsang, W. H. H., & V. Pearson (1996). A conceptual framework on work-related social skills for psychiatric rehabilitation. Journal of Rehabilitation. 62(3), 61-67.

Tsang, W. H. H., & Pearson, V. (a) (in press). A work-related social skills training for people with schizophrenia in Hong Kong. Schizophrenia Bulletin.

Tsang, W. H. H., & Pearson, V. (b) (in press). Reliability and validity of a simple measure for assessing the social skills necessary for seeking and securing a job of people with schizophrenia. Canadian Journal of Occupational Therapy.

Unger, K. V., & Anthony, W. A. (1984). Are families satisfied with services to young adult chronic patients. A recent survey and a proposed laternative, In B. Pepper & H. Ryglewicz (Eds.), Advances in treating the young adult chronic patients (New Directions for Mental Health Services, No. 21, pp. 91-97). San Francisco: Jossey-Bass.

Vaillant, G. E. (1964). Prospective prediction of schizophernic remission. Archives of General Psychiatry, 11, 509-518.

Van OS, J., Fahy, T. A., Jones, P., Harvey, I., Lewis, S., Williams, M., Toone, B., & Murray, R. (1995). Increased intracerebral cerebrospinal fluid spaces predict unemployment and negative symptoms in psychotic illness: A prospective study. British Journal of Psychiatry, 166, 750-758.

Vetter, P., & Koller, O. (1996). Clinical and psychosocial variables in different diagnostic groups: Their interrelationships and value as predictors of course and outcome during a 14-year follow-up. Psychopathology, 29, 159-168.

Vogel, R., Bell, C., Blumenthal, St., Neumann, N. U., & Schuttler, R. (1988). Outcome, course, and prognosis of the employment situation of first-admission psychiatric patients: Findings of a multiple-point investigation. Rehabilitation, 27, 5-13.

Wasylenki, D. A., Goering, P. N., Lancee, W. J., Ballantyne, R., & Farkas, M. (1985). Impact of a case manager program on psychiatric aftercare. The Journal of Nervous Mental Disease, 173, 303-308.

Wohrl, H. G. (1990). Persons with chronic mental illness: Their occupational integration outlook following vocational retraining at the BFW Heidelberg. Rehabilitation, 29, 84-92.

Hector Tsang Paul Lam The Hong Kong Polytechnic University

Bacon Ng Castle Peak Hospital

Odelia Leung Kwai Chung Hospital

Hector Tsang, Ph.D., Assistant Professor, Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hunghom, Hong Kong. Email: [e-mail]