Objective: This study examined whether information obtained early in the ho
spitalization process can be used to assess a patients need for extended ca
re. Methods: A sample of 2,430 inpatients who were admitted to a state psyc
hiatric facility during a one-year index period (January through December 1
997) were randomly assigned to a primary sample or a replication sample. Da
ta were collected on demographic characteristics and history of previous ho
spitalization. The Brief Psychiatric Rating Scale-Anchored Version (BPRS-A)
was administered to patients within 48 hours of admission, and four new su
bscales derived from ratings of newly admitted patients were calculated. Un
ivariate and multivariate analyses were conducted to identify factors assoc
iated with whether a patient was discharged to the community or transported
to another hospital for extended care. Results: A discriminant analysis of
the data correctly identified 70 percent of the patients who were referred
for continued hospitalization and 80 percent of the patients who were disc
harged to the community. The main correlates of the need for extended inpat
ient services were, in descending order, scores on the BPRS-A resistance su
bscale, the number of previous referrals for extended hospitalizations, and
scores on the BPRS-A positive symptoms and psychological discomfort scales
. Conclusions: BPRS-A subscale scores should be considered to be at least a
s good as more traditional measures in predicting length of hospitalization
.