This paper is one of a series reporting on a clinical field trial evaluatin
g the efficacy of the modified therapeutic community (TC) approach for the
treatment of homeless mentally in chemical abusers (MICAs). The social and
psychological characteristics of the treatment sample were described in an
earlier paper; the purpose of the present report was to categorize subtypes
of homeless MICA clients to predict with greater accuracy their treatabili
ty in modified TCs. An index that consistently correlated with treatment-re
levant variables was identified for each of three dimensions; Homelessness
(residential instability), Mental Illness (current severity, and Substance
Abuse (current substance abuse/dependence diagnosis). These indices yielded
distributions that captured the variability in the sample with respect to
a number of variables, including drug use, criminality, human immunodeficie
ncy virus (HIV) risk (sexual behavior), psychological status, and motivatio
n. Bivariate and multivariate analyses showed that the indices were not str
ongly related to demographic variables such as race/ethnicity, age, or gend
er, but were significantly associated with baseline drug use, criminal acti
vity, HIV risk (sexual behavior), psychological symptoms, and motivation an
d readiness. These findings indicate that, even among those admitted to res
idential treatment for substance abuse, homeless MICA clients are not homog
eneous; rather, subgroup differences emerge among the indices of homelessne
ss, mental illness, and substance abuse. The efficacy of treatment in modif
ied TCs for these subgroups will be assessed in subsequent papers examining
the relationships among the three indices, client retention, and outcomes
during and subsequent to residential treatment.