Table 1.

Main variables used in the analysis

Service types

Medical services. These include medical emergency care visits, medical inpatient and outpatient care, use of medical supplies, and pharmaceutical services. Psychiatric services. These are services related to assessment and treatment of psychiatric illness, including psychiatric inpatient, outpatient, and emergency care, psychiatric day treatment programs, and psychopharmaceutical services. Specific components of these services include individual and group psychotherapy, psychotropic medication management, and psychiatric case management. Substance abuse services. These services represent substance abuse treatment including inpatient, residential, and outpatient care. Specific components of these services include individual and group psychotherapy for substance abuse and psychotropic medication management for substance abuse as well as self-help groups such as Alcoholics Anonymous and Narcotics Anonymous. Homeless maintenance services. These services relate to the maintenance of individuals while they are homeless, including food (e.g., soup kitchens and food pantries), transportation services, and public and temporary housing services such as nighttime sheltering and transitional housing. Homeless amelioration services. These services relate to the amelioration of homelessness including housing services, vocational rehabilitation, and legal aid services. Total services. These services represent the sum of the five types of service variables above. Acute mental health services. These services refer to acute psychiatric and substance abuse services, including emergency and inpatient care for psychiatric and substance use disorders. Outpatient mental health services. These services refer to psychiatric and substance abuse services provided in non-emergent outpatient settings.

Housing status

Housed during Year 1. Report of having stable housing for most of the first follow-up year and for the last 30 days of that year.

Housed during Year 2. Regardless of housing status in the first year, stable housing was reportedly obtained for most of the second follow-up year and for the last 30 days of that year.

Combined Years 1 and 2 housed. Housed status in the first year was concatenated with housed status in the second year to create a variable representing all housed years. (Note that some of the individuals housed in the second year were also housed in the first year and others were not housed in the first year. Hence, this variable does not represent individuals. It represents the total number of housed years among the members of the sample.)

Combined Years 1 and 2 unhoused. Unhoused status in the first year was concatenated with unhoused status in the second year to create a variable representing all unhoused years. (Again, this variable represents the total number of unhoused years, not individuals.)