Prediction
of remission from depression by responses on two
items of HAM-D17 rating
scale
Evidence
from studies indicates that the Hamilton
Depression (HAM-D) rating scale is relatively
insensitive to improvements in some depressive
symptoms which increases the likehood of falsely
classifying patients as non-responders. The use
of a semi-standardized interview and subscales
derived from the HAM-D improves the detection of
response and remission after antidepressant
treatment. Recently, two items from the
HAM-D17 scale, depressed mood (item
1) and psychic anxiety (item 10) have been
demonstrated to have enhanced sensitivity for
detecting improvement in a study comparing the
antidepressant efficacy of the serotonin and
noradrenaline reuptake inhibitor, venlafaxine,
and placebo.
More recently the responses on these items 1 and
10 were studied following treatment with
venlafaxine or a selective serotonin reuptake
inhibitor, to determine the degree of prediction
of remission from depression.
The subjects (2027 outpatients), who met the
DSM-III-R/-IV criteria for major depressive
disorder, were enrolled into eight active- and
placebo-controlled studies to receive
venlafaxine, fluoxetine, paroxetine, fluvoxamine
or placebo for 6 to 8 weeks. Efficacy
assessments were made on study days, 1, 7, 14,
21, 28, 42, and 56 with items 1 and 10 scores
taken from the HAM-D17 scale. Item 1
and 10 scores were combined to create an Absence
of Depressive and Anxious Mood (ADAM) score. The
patients with a score of zero for the depressed
mood item and psychic anxiety item were
designated as responders.
Data analysis showed that venlafaxine improved
depressed mood, psychic anxiety and ADAM scores
after 2 weeks treatment more effictively than
SSRIs or placebo. In addition, the odds ratios
for the ADAM indicated a high specificity for
predicting a clinical remission defined as total
HAM-D17 score </= 7.
These findings demonstrate that these two HAM-D
items may be sufficient to assess treatment
response and to predict remission rates.
Therefore, in contrast to outcomes from total
HAM-D scores, commonly used in clinical trials
assessing depression severity, which are not
very sensitive, unidimensional subscales may
represent a more reliable and finally a more
attractive predictive value for clinicians. Int
Clin Psychopharmacol 17: 1-9, 2002