Measures and predictors of community-based employment and earnings of persons with schizophrenia in a multisite study.

OBJECTIVE: Data from a national study of persons with schizophrenia-related disorders were examined to determine clinical factors and labor-market conditions related to employment outcomes. METHODS: Data were obtained from the U.S. Schizophrenia Care and Assessment Program, a naturalistic study of more than 2,300 persons from organized care systems in six U.S. regions. Data were collected via surveys and from medical records and clinical assessments at baseline and for three years. Outcome measures included any community-based (nonsheltered) employment, 40 or more hours of work in the past month, employment at or above the federal minimum wage, days and hours of work, and earnings. Bivariate and multiple regression analyses of data from more than 7,000 assessments tested relationships between outcomes and sociodemographic, clinical, and local labor market characteristics. RESULTS: The employment rate was 17.2%; only 57.1% of participants who worked reported 40 or more hours of past-month employment. The mean hourly wage was $7.05, and mean monthly earnings were $494.20. Employment rates and number of hours worked were substantially below those found in household surveys or in baseline data from trials of employment programs but substantially higher than those found in a recent large clinical trial. Strong positive relationships were found between clinical factors and work outcomes, but evidence of a relationship between local unemployment rates and outcomes was weak. CONCLUSIONS: Work attachment and earnings were substantially lower than in previous survey data, not very sensitive to labor market conditions, and strongly related to clinical status.

Prevalence, severity, and co-occurrence of chronic physical health problems of persons with serious mental illness.

OBJECTIVES: This study examined Medicaid claims forms to determine the prevalence, severity, and co-occurrence of physical illness within a representative sample of persons with serious mental illness (N=147). METHODS: Representativeness of health problems in the study sample was established through comparison with a larger sample of persons with serious mental illness enrolled in Medicaid within the same state. Standardized annual costs were then assigned to Medicaid claims diagnoses, and individual health problem severity was measured as the sum of estimated treatment costs for diagnosed conditions. RESULTS: Seventy-four percent of the study sample (N=109) had been given a diagnosis of at least one chronic health problem, and 50 percent (N=73) had been given a diagnosis of two or more chronic health problems. Of the 14 chronic health conditions surveyed, chronic pulmonary illness was the most prevalent (31 percent incidence) and the most comorbid. Persons with chronic pulmonary illness were second only to those with infectious diseases in average annual cost of treatment ($8,277). Also, 50 percent or more of participants in eight other diagnostic categories had chronic pulmonary illness. A regression analysis identified age, obesity, and substance use disorders as significant predictors of individual health problem severity. CONCLUSIONS: Risk adjustment for physical health is essential when setting performance standards or cost expectations for mental health treatment. Excluding persons with chronic health problems from mental health service evaluations restricts generalizability of research findings and may promote interventions that are inappropriate for many persons with serious mental illness.