In a three phase sequential treatment strategy study involving 119 dep
ressed outpatients, a total of 31 patients (26.7%) stopped treatment p
rematurely due to side-effects (21/31), aggravation of symptoms (3/31)
; non-compliance (4/31), and non-treatment related events (3/31). At b
aseline, no significant differences were found on sociodemographic and
psychiatric data between patients who did or did nor drop out. As pre
dictors of dropout eight domains of data concerning psychiatric histor
y, premorbid history, symptomatology, personality, and social adjustme
nt were used. Using a logistic linear regression analysis, only three
variables were related to dropout. Patients with a history of alcohol
use or poor social functioning according to axis V of the Diagnostic a
nd Statistical Manual (DSM)-III-R had a higher chance to drop out, whi
le patients with a sleep disturbance according to the Symptom Checklis
t (SCL)-90 had a smaller chance to drop out. (C) 1998, Elsevier, Paris
.