J. Neurol. Neurosurg. Psychiat., 1960, 23, 56.

A RATING SCALE FOR DEPRESSION

BY
MAX HAMILTON
From the Department of Psychiatry, University of Leeds

The appearance of yet another rating scale for appear in different settings. Other symptoms are
measuring symptoms of mental disorder may seem difficult to define, except in terms of their settings,
unnecessary, since there are so many already in e.g., mild agitation and derealization. A more
existence and many of them have been extensively serious difficulty lies in the fallacy of naming. For
used. Unfortunately, it cannot be said that per- example, the term "delusions" covers schizophrenic,
fection has been achieved, and indeed, there is depressive, hypochrondriacal, and paranoid de-
considerable room for improvement. lusions. They are all quite different and should be
clearly distinguished. Another difficulty may be
Types of Rating Scale summarized by saying that the weights given to
The value of this one, and its limitations, can best symptoms should not be linear. Thus, in schizo-
be considered against its background, so it is useful phrenia, the amount of anxiety is of no importance,
to consider the limitations of the various rating whereas in anxiety states it is fundamental. Again,
scales extant. They can be classified into four a schizophrenic patient who has delusions is not
groups, the first of which has been devised for use necessarily worse than one who has not, but a
on normal subjects. Patients suffering from mental depressive patient who has, is much worse. Finally,
disorders score very highly on some of the variables although rating scales are not used for making a
and these high scores serve as a measure of their diagnosis, they should have some relation to it.
illness. Such scales can be very useful, but have Thus the schizophrenic patients should have a high
two defects: many symptoms are not found in score on schizophrenia and comparatively small
normal persons; and less obviously, but more scores on other syndromes. In practice, this does
important, there is a qualitative difference between not occur.
symptoms of mental illness and normal variations The present scale has been devised for use only on
of behaviour. The difference between the two is not patients already diagnosed as suffering from
a philosophical problem but a biological one. There affective disorder of depressive type. It is used for
is always a loss of function in illness, with impaired quantifying the results of an interview, and its value
efficiency. depends entirely on the skill of the interviewer in
Self-rating scales are popular because they are eliciting the necessary information. The interviewer
easy to administer. Aside from the notorious un- may, and should, use all information available to
reliability of self-assessment, such scales are of little help him with his interview and in making the final
use for semiliterate patients and are no use for assessment. The scale has undergone a number of
seriously ill patients who are unable to deal with changes since it was first tried out, and although
them. there is room for further improvement, it will be
Many rating scales for behaviour have been found efficient and simple in use. It has been found
devised for assessing the social adjustment of to be of great practical value in assessing results of
patients and their behaviour in the hospital ward. treatment.
They are very useful for their purpose but give little
or no information about symptoms. Description of the Rating Scale
Finally, a number of scales have been devised The scale contains 17 variables (see Appendix I).
specifically for rating symptoms of mental illness. Some are defined in terms of a series of categories
They cover the whole range of symptoms, but such of increasing intensity, while others are defined by a
all-inclusiveness has its disadvantages. In the first number of equal-valued terms (see Appendix II).
place, it is extremely difficult to differentiate some The form on which ratings are recorded also includes
symptoms, e.g., apathy, retardation, stupor. These four additional variables: Diurnal variation, de-
three look alike, but they are quite different and realization, paranoid symptoms, obsessional symp-
56
 A RATING SCALE FOR DEPRESSION 57
toms. These are excluded from the scale because the extremely difficult to get a satisfactory description of
first is not a measure of depression or of its intensity, them from the patient.
but defines the type of depression. The other three Hypochondriasis.-This is easy to rate when it is
occur so infrequently that there is no point in obviously present, but difficulties arise with mild
including them. hypochondriacal preoccupations. Phobias of spe-
The variables are measured either on five-point cific disease can cause difficulties. A phobia of
or three-point scales, the latter being used where venereal disease or of cancer will sometimes be
quantification of the variable is either difficult or rated under "guilt" by the nature of the symptom,
impossible. No distinction is made between in- but other cases may give rise to much doubt and
tensity and frequency of symptom, the rater having judgment requires care. Fortunately, phobias are
to give due weight to both of them in making his not common, but the whole subject of hypochon-
judgment. driasis could well repay clinical investigation.
Various problems are to be found with specific
symptoms. Thus considerable difficulty is found Insight.-This must always be considered in
with the depressive triad: depressive mood, guilt, relation to the patient's thinking and background
and suicidal tendencies. These are so closely linked of knowledge. It is important to distinguish be-
in description and judgment as to be very difficult to tween a patient who has no insight and one who is
separate. It is very important to avoid the halo effect reluctant to admit that he is "mental".
by automatically giving all of them high or low Loss of Weight.-Ideally this would be measured
scores, as the case may be. in pounds or kilograms, but few patients know their
Depressed Mood.-This tends to have a narrow normal weight and keep a check on it. It was
range of scores, for no diagnosed patients will score therefore necessary to use a three-point scale.
zero and few will score 1 or 4. The most useful After recovery from depression, some patients
indicator for depressed mood is the tendency to sometimes show a brief hypomanic reaction, during
weep, but it must always be considered against the which the exuberantly cheerful patient will deny that
cultural background, and patients may also "go he has any symptoms whatever, though he is ob-
beyond weeping". viously not to be regarded as normal. In such cases,
Suicide.-An attempt at suicide scores 4, but such the rating scale is inapplicable and should be
an attempt may sometimes occur suddenly against a delayed until the patient has fully recovered.
background of very little suicidal tendency; in such
cases it should be scored as 3. There will be great Scoring
difficulty sometimes in differentiating between a It is particularly useful to have two raters in-
real attempt at suicide and a demonstrative attempt; dependently scoring a patient at the same interview,
the rater must use his judgment. since this gives data for calculating the inter-
Work and Loss of Interest.-Difficulties at work physician reliability. The score for the patient is
and loss of interest in hobbies and social activities obtained by summing the scores of the two
are both included. The patient who has given up physicians. This is, of course, the best way of
work solely because of his illness is rated 4. learning how to use the scale, Where only one rater
uses the scale, the scores should be doubled so as
Retardation.-A grade 4 patient is completely to make them comparable. With sufficient ex-
mute, and is therefore unsuitable for rating on the perience, a skilled rater can learn to give half-points.
scale. Grade 3 patients need much care and patience
to rate, but it can be done.
Results
Agitation.-This is defined as restlessness asso- For two raters, the correlation between summed
ciated with anxiety. Unfortunately, a five-point
scale was found impracticable, and therefore this scores for the10 first 10 patients was 0-84. Adding
variable is rated on a three-point scale. The mildest successively patients at a time, the correlation
degrees of agitation cause considerable difficulty. changed to 0X84, 0-88, 0-89, 0-89, 0 90, 0 90. The
last correlation is therefore total for 70 patients.
Gastro-intestinal Symptoms.-These occur in con- Product-moment correlations were calculated for
nexion with both anxiety and depression. Con- the 17 variables on the first 49 male patients
siderable clinical experience is required to evaluate (Table I). The correlation matrix was then factor-
them satisfactorily. The definitions given have been analysed by extracting the latent roots and vectors
found very useful in practice. (Table II). As the intercorrelations are in general
General Somatic Symptoms.-In depressions these low because of the intense selection of patients, the
are characteristically vague and ill defined, and it is latent roots (variances extracted by factors) diminish