Lecture 2
CLASSIFICATION
Lecture Outline
I. Introduction
II. Historical Systems
III.General Problems
IV. The Diagnostic and Statistical Manuals
A. DSM I
B. DSM II
C. DSM III
D. DSM III-R
V. General Classification Issues
A. Natural Kinds
B. Language
VI. Diagnosis and Classification
------------------------------------------
I. Introduction
We have discussed some of the issues involved in defining "abnormal
behavior" per se. But clearly, "defining" does not stop there. Even
the lay person is aware that there exist various types of abnormal
behaviors. The task for the psychologist, then, is to specify what
these different types are. Some classification system is necessary.
Botanists would never have advanced their field without a clear taxonomy
for classifying plants. Mere "Plant" - "Not Plant" distinctions would
not be very fruitful! So, too, we need to have a system that allows us
to identify the different types of disorders which fall under the rubric
of "Abnormality".
Importance of classification (see, for example, Blashfield &
Draguns, 1976):
A. Description - identification of presence of a disorder.
B. Communication - need a vocabulary.
C. Research - homogeneous groups.
D. Treatment decisions - intervention, prognosis.
E. Third-party reimbursement - insurance companies require a
diagnosis.
F. Funding for programs.
G. Theory development.
In striving for a classification system, a number of things have
traditionally been sought (see, for example, Sprock & Blashfield, 1983):
A. Exhaustive system: the classification system should account
for all the types of abnormal behavior that we encounter. A
classification system for plants would be of limited use if it
had nothing to say about all the plants in your backyard.
B. Mutually exclusive categories/Independent dimensions: Your
plant classification system should clearly distinguish between
plants, even those that are very similar. There should be no
doubt which plant this plant is. Likewise, the most powerful
classification system for mental disorders will not result in
fuzzy categories and uncertainty over whether this person
suffers from disorder A or disorder B. [Actually, not all
classification systems rely on distinct categories - some base
classification on determining where a person falls on various
dimensions -all people falling somewhere on each of the
dimensions (eg: some possible dimensions might be
"emotionality" or "introversion"). These dimensions, however,
must be independent. If two or more dimensions overlap, then
they're not really measuring separate aspects of a person.
Just like we wanted our categories to be separate, in this
approach, the different dimensions need to be independent.
We'll have more to say about this approach later in the
course].
C. Reliability: Each time you use your classification system,
it should give you the same result if you're classifying the
same individual. If yesterday I classified this plant as a
rose based on classification system X, then using the same
system X, I (or anyone else) will today classify that same
plant as a rose. A classification system that leads me to
diagnose this person as schizophrenic today and depressed next
week isn't very useful. [Note: Reliability suffers when
conditions (A) and (B) above are not met].
D. Validity: Finally, a classification system should be
saying something that is true about the "real world". If my
classification system is supposed to be classifying mental
disorders, than that is what it should do! If my system has
categories such as schizophrenia, anxiety disorder, and
depression, then there should actually exist these disorders.
No matter how nice it would be to have the "Ronald Reagan
Bush" in my plant classification system, a system with such a
category would not be completely valid. In fact, a whole
system might be invalid: My system categorizes people
according to shoe size - those with sizes 0-3 are depressed,
4-7 are anxious, 8-10 are schizophrenic, 11-13 suffer from
borderline personality disorder. Note, this is a very
reliable system (shoe size can be measured very accurately),
but it is totally invalid!
II. Historical systems
A. Probably began in ancient pre-history (indeed, probably even
with pre-humans), when people first noticed differences in others
and wanted to understand those differences.
B. Ancient Greece: Hippocrates, fourth century B.C. Elaborated a
medical model of madness, centering on the interactions of four
bodily fluids ("humors") - blood, black bile, yellow bile, and
phlegm - and devised a classification system based on the four
corresponding temperaments - sanguine, melancholic, choleric, and
phlegmatic . The specifics aren't important here (although make
sure you review Hippocrates' theory as discussed in your text, pp.
31-32); what is important is that Hippocrates and his followers
rejected the prevailing notion that illness was of divine or sacred
origin. Thus, as physical illness could be classified, so too
could mental illness. Hippocrates' approach, if not his actual
theory, was the foundation for classification systems for centuries
to come (Veith, 1957).
C. Other developments: (Kaplan & Sadock, 1981; Veith, 1957).
1. Jean Fernel (1497-1588): An Anatomical Approach - related
disease and body structure. First to use the word "physiology"
in the modern sense, and coined the word "pathology". [Some
modern researchers have followed in Fernel's footsteps. Sheldon
(1954), for example, tied physical body "types" to personality
characteristics. There is little evidence, however, to support
this approach].
2. Felix Platter (1536-1614): A new classification system of
illnesses, based on symptomatology.
3. Late 17th Century: Physicians and neurologists engaged in a
giant effort to create an all-encompassing taxonomic system,
with irreducible and constant units. Hopes are high.
4. Francois Baussier de Sauvages (18th Century): developed a
system based on his detailed observations. Consisted of 10
classes, 40 orders, 78 genera, 2,400 distinct diseases! The
problem with this system is that it failed to differentiate
between symptoms and diseases:
"symptom" = an observable (physiological, psychological,
etc) manifestation of a disease.
"disease" = an impairment of the normal state of the
organism, identified by the particular group of symptoms
and a specific cause.
5. Philippe Pinel (1745-1826): French psychiatrist and a pupil
of Sauvages. Developed the first workable psychiatric
nosography. 5 forms of insanity: mania, melancholia with
delirium, melancholia without delirium, dementia, idiotism. In
this system, Pinel avoided physiological fictions and
metaphysical hypotheses. His system was purely descriptive.
6. Emil Kraepelin (late 19th century): A student of Wilhelm
Wundt, and the "father of systematic psychiatry" (Weckowicz,
1984). Kraepelin's goal was to devise an accurate definition of
the separate disease processes and disease entities. He made a
vast number of observations, collected case histories on
hundreds of people, and eventually published his two volume,
2,425 page Lehrbuch. This book contains his medical
classification scheme, the first truly comprehensive system. In
it he defines two major groups of mental disorders: the manic-
depressive psychoses, and dementia praecox (ie schizophrenia).
These were further subdivided into 18 specific mental disorders.
Kraepelin's system is important because it defined the
discipline of psychiatry as distinct from neurology, and as a
part of the general field of medicine. Kraepelin's influence is
felt even in many modern classification systems.
III. General problem with these classification systems
A. Unreliable: Consistent diagnoses were difficult to obtain. In
addition, specific characteristics of a disorder as
described in the various classification systems were
inconsistently associated with the disorder they
were supposed to be associated with.
B. Often theory laden: Often the only support for a particular
classification system was theoretical - there was no
actual evidence/data to support the system (or the
theory!). In addition, the theory ladenness of
these systems made it virtually impossible for
practitioners of different theoretical persuasions
to communicate with each other. As a result, some
people argued that the ideal classification system
should be theory-free and thereby useful to all.
C. Overlapping categories: It was difficult to distinguish between
one category and another - difficult to know where
to "place" someone into the diagnostic system.
D. Lack validity: There was little evidence that the nature of
psychological disorders was accurately represented
by these classification systems.
IV. The Diagnostic and Statistical Manuals
A. DSM-I: The American Psychiatric Association, in the 1950's,
attempted to address these problems. In 1952 they published DSM-I,
the next step in the evolution of psychiatric taxonomy.
Nevertheless, it was still largely unreliable and some complained
that it was too theory laden, especially with a psychodynamic view
(APA, 1987). Note: It is not altogether clear why a
classification should be (or even could be) theory-free...but more
on this in the future.
B. DSM-II: A goal of this 1968 revision was compatibility with the
World Health Organization's International Classification of
Diseases, 8th ed. (ICD-8) which covers all diseases and disorders,
both physical and psychological, for the entire world (it, too, has
since been revised and updated). The APA and WHO worked closely,
consulting with each other to ensure compatibility. In addition,
DSM-II attempted to use term that did not imply any particular
theoretical framework. The goal was to be purely descriptive.
C. DSM-III: But there were still problems with the second edition:
continuing unreliability, still largely psychodynamic in
orientation, incompatible or inconsistent with new research
advances and new diagnostic methodologies (tests, interviews,
etc.). The ICD was of little help - the ICD was not sufficiently
detailed for clinical and research use in the U.S.. The APA was
striving for a system that would be as clinically useful as
possible to a wide range of professionals (APA, 1987).
During the 70's, a number of efforts were made to address these
problems. Attempts were made to be as atheoretical as possible and
instead rely on purely descriptive terminology. As a result,
reliability indeed improved with these new classification systems.
In 1972, Feighner and his coworkers (Feighner, Robins, Guze, et
al., 1972) published a classification system which established
explicit definitions for 16 psychiatric disorders. There were
clearly defined rules about which characteristics were necessary
and how many were sufficient to make a particular diagnosis.
Spitzer, Endicott, and Robins built on Feighner's work by adding
criteria regarding the severity, duration, and course of specific
disorders. Their work resulted (in 1978) in the publication of the
Research Diagnostic Criteria (RDC), containing explicit diagnostic
criteria for 25 diagnostic categories.
Based on the work of people like Feighner and Spitzer, in 1980 the
APA released the DSM-III. This was a "bold departure" from the
previous DSM's (Smith & Kraft, 1982). The DSM-III has a number of
important characteristics (APA, 1980):
1. Descriptive approach, largely because etiology is typically
unknown, debatable, or theory laden. Thus, DSM-III strives to
be atheoretical, using merely observationally referent terms.
The hope with this is to make the manual as acceptable as
possible to professionals with different theoretical
orientations.
2. Specific criteria for making diagnoses (not provided in I,
II, or the ICD).
3. Systematic description: the DSM-III provides information on
essential features, associated features, prevalence rates, sex
ratios, family patterns, and differential diagnoses (ie:
possible alternative or additional diagnoses that should be
considered).
4. Multiaxial system: this is the major advance over II. The
multiaxial approach (to be described below) allows for a more
holistic and comprehensive account of an individual (Smith &
Kraft, 1983).
D. DSM-III-R: In 1987, the APA published a revised version of DSM-
III. The goals of this revision (APA, 1987) were:
1. expanded coverage
2. increased reliability (a major complaint with the 3rd
edition was the unreliability of the personality disorders)
3. updated with the latest research findings
Note: It is very difficult to do both 1 and 2 at the same time.
The more coverage a system has, the more possibility for differing
diagnoses, and hence unreliability. On the other hand, the fewer
the number of categories, the higher will be a system's
reliability, but the increased likelihood that individuals will not
fit perfectly into any particular category. The goal of DSM-III-R
is to avoid such mixed or atypical diagnoses, while at the same
time have good coverage and reliability (Morey, 1988).
Thus, the revised manual contain over 100 changes in its diagnostic
criteria (Peele, 1986), and also changes in its multiaxial system
of diagnosis.
DSM-III-R Multiaxial System
A person is assessed on several different axes or dimensions, each
focusing on a different type of information.
Axis I Clinical Syndromes
Axis II Developmental disorders and Personality disorders
Axis III Physical disorders and conditions
Axis IV Severity of psychosocial stressors
Axis V Global assessment of functioning
I and II comprise the entire classification of mental disorders,
plus "V codes" (codes which indicate conditions not attributable to
a mental disorder but that are a focus of attention or treatment.
Eg: Academic, interpersonal or occupational problems).
I, II and III together constitute the official, complete DSM-III-R
diagnostic assessment.
I: Clinical syndromes: Florid psychological disorders, such
as major depression or panic disorder.
II: Developmental and Personality disorders: Pervasive, long
standing disorders, typically beginning in childhood or
adolescence. Eg: Mental retardation or borderline personality
disorder. A person can have a diagnosis on both axis I and
II.
III: Physical disorders/conditions: Current physical
complaints that may be relevant to understanding or managing
the case. Eg: Neurologic disorders or diabetes.
IV: Severity of psychosocial stressors: Overall severity of
life stress for the past year. (Stress is clearly related to
mental health, including the development, recurrence, and
exacerbation of a disorder).
Types of stressors:
marital financial
parenting legal
interpersonal phases of the life cycle
occupational physical illness/injury
living circumstances disasters, rape, persecution...
These stressors are rated on a six point scale, ranging from
"None" to "Catastrophic".
V. Global Assessment of functioning: This allows the
clinician to give his/her judgment of the person's
psychological, social and occupational functioning for two
time periods:
1. Current: reflects need for treatment
2. Past Year-highest level of functioning: reflects
possible prognosis
Each is rated on a 90 point scale, ranging from 1 (Suicidal
acts, recurrent violence, etc) to 90 (Absent or minimal
symptoms). [Note: In the DSM-III system, Axis V referred to
"Highest Level of Adaptive Functioning Past Year" - social and
occupational functioning for the past year only. The DSM-III-
R expands this to include social, occupational as well as
psychological functioning for both the past year and the
present. In addition, the DSM-III rated functioning on just a
7 point scale ("superior" to "grossly impaired")].
V. General Classification Issues
A. Natural kinds: The assumption behind diagnostic systems like
DSM-III-R is that illness is a "thing in itself" (Veith, 1957) -
that each disorder is a distinct entity. But are there "natural
kinds"? That is, are there actually disorders that truly exist
"out there" in the world. Or are the disorders as defined in the
DSM-III-R simply figments of scientists' imaginations! We're
talking about "reality" (what ever that is!) -and whether or not
the DSM-III-R has accurately "cut it up". The DSM is assuming that
the disorders it defines are real and accurately defined. There is
a controversy, to which we will return to in detail later in the
course, over whether the various mental illnesses (however defined:
by the DSM-III-R or by any other system) are social constructions
rather than actual "things" (cf Wartofsky, 1983: "childhood" as a
cultural invention). That is, schizophrenia, for example, doesn't
really exist as a real disease (the way, say, diabetes is real),
but is at best a hypothetical entity which happens to help us
organize and simplify our thinking about some complex human
behavior (Suppe, 1977). Some have gone so far as to claim that
mental illness is entirely a myth - that there is no such thing at
all and it is actually harmful to think that there is (Sharman,
1970; Szasz, 1961). But, again, more on this later in the course.
At this point, I just want you to get your "critical caps" placed
firmly on your heads - We'll be talking about all sorts of
disorders in this course (schizophrenia, major depression,
agoraphobia...), but how do we know what these things really are?
Do they exist? How do we define them? Is there a better way to
think about abnormal behavior? How would you know it's a better
way?
B. Inadequate language of description (Rotter, 1954): Our terms,
even purely descriptive ones, have too many referents (ie: a single
word can refer to many different things, depending on how it is
used), unclear referents, or nonspecific referents. This can lead
to Fuzziness: Words used in clinical psychology are often vague.
Intelligence, Frustration, Anxiety and Depression, for example,
have no clear referents. These terms can also have elastic meaning
depending on one's perspective. So, it's difficult to know with
precision what we are talking about! When I say "anxiety", what
precisely do you hear or understand? Do each of you think of
exactly the same thing? Unlikely.
C. Reification: A particularly difficult problem is the problem of
reification. The words we use to describe people, such as
"anxious", "depressed", and so on were meant for just that: to
describe them. However, there is the danger that those terms will
be turned around to explain people's behavior: "Why does she act
that way?" "Because she is depressed". The description becomes the
explanation - a circular and thus meaningless explanation is set
up. The description is used to explain itself! Unfortunately,
this is a very common problem, not just with our everyday thinking,
but also in scientific thinking. Let's keep our eyes open for it!
The DSM-III-R tries to address at least the "fuzziness" issue. Whatever
its shortcomings, it is important to understand that the DSM "freezes in
time `one still frame' in the area of study that is constantly
advancing" (Smith & Kraft, 1983).
VI. A note on Diagnosis and Classification
Classification is a central aspect of the overall diagnostic
process (Cooper, 1983).
Diagnostic process:
1. Decisions about what data to collect.
2. Data collection - interviews, tests, observations of
behavior.
3. Sort data and select aspects of it to make inferences and
decisions regarding symptoms, traits, disabilities, skill
deficits, etc.
4. Summarization of this potentially diagnostic information.
5. Use information to place client into a particular
classification (ie: give a diagnosis).
The diagnostic process is itself an aspect of the final clinical
formulation or assessment. This will include: diagnosis, history,
current situation, interpersonal and social context. The diagnosis is a
general indicator of the type of problem(s) to be expected, the tendency
toward certain behavioral and emotional patterns.