Transcript The Measurement of Mental Disorder

The Measurement of Mental
Disorder
Why is it so difficult to determine who
is mentally disordered when
interviewing people in the community?
1
Why do community sampling?
 Helpful in formulating mental health policy
 Helpful in evaluating theories of causation
 Helpful in planning delivery of mental health
services
 Helpful in justifying requests for funding to
support research and service provision
2
2 Major Epidemiologic Studies
 Epidemiologic Catchment Area (ECA) Project


Robins & Regier, 1991
Surveyed samples of general population in 5 sites—New
Haven, CT to Los Angeles, CA
 National Comorbidity Survey (NCS)


Kessler et al., 1994
Focused on degree to which different disorders appeared
in same individual
3
Difficulties Encountered
 Prevalence rates of mental disorders seem to be too
high


ECA found that 1/3 of Americans suffer from a mental
disorder at one point in their lives
NCS figure was 1/2
 Although the studies used similar instruments, the
results were quite different in many areas


ECA estimated that 6% of Americans suffer from a
depressive disorder at some point in their lives
NCS estimated 17%
4
Clinical prevalence vs. true
prevalence
 Why not just count the number of people in clinical
settings?
 Clinical prevalence of mental disorder is not a
measure of true prevalence

Clinical prevalence is a measure of how many people
seek and receive treatment; true prevalence is a measure
of how many people have mental disorders, whether or
not they seek or receive treatment
 It’s the rate of untreated mental disorder that is
important for planning purposes.
5
What does socioeconomic status (SES)
have to do with mental disorder?
 Diverse mix of services may not be accessible, especially in
lower-class communities


In lower class communities, there are fewer office therapists
available (poor people can’t afford them)
Poor people more likely to use public hospitals; middle and upperclass people more likely to use office therapists
 Social class, ethnicity, and gender have an effect on how
mental disorders are expressed, diagnosed, and handled
 Lower SES people may be under more chronic stress, may
be more prone to develop mental disorders
6
Conceptual validity
 Measurement tool has conceptual validity when it is
successful in distinguishing between persons who
have a mental disorder and those who don’t
 Too expensive to have clinicians interviewing large
community samples, so questionnaires are used. It
is very difficult to create a questionnaire that has
conceptual validity.
7
Back to the DSM
 Basic idea is that disorders can be recognized by
symptoms
 DSM gives rules for deciding, from symptoms,
which disorder a person has, if any
 DSM is theory-neutral
 DSM is useful tool for psychiatric epidemiologists
8
Fundamental question
 What is mental disorder, and how does it
differ from the “normal suffering of everyday
life” of “problems in living”?
 According to DSM, pattern of symptoms that
is identified must be associated with distress,
disability, dysfunction. Dysfunction must be
harmful. Disturbance must not be only a
conflict between an individual and society.
9
Validity
 Criteria for classifying a disorder must be:


Sensitive: when applied the criteria should not
indicate that someone is normal when, in fact,
they have a disorder (false negative)
Specific: when applied the criteria should not
indicate that someone has a disorder when they
are normal (false positive)
10
More on validity
 False positives more of a problem than false
negatives—because the normal population is
much larger than the disordered population.
 If there are many false positives, the
estimated prevalence of a disorder can be
very much overestimated.
11
Problem with using DSM
criteria in epidemiology
 DSM is designed to be used in a clinical
setting, where clinicians have the luxury of
refining the diagnosis after several
interviews.
 Questionnaires used in community surveys
give epidemiologist a quick diagnosis, one
that cannot be refined later.
12
Problems with epidemiological
studies of mental disorders
 No matter how carefully the instrument for
measurement is developed, sample is
selected, and data is collected, if the
definitions of mental disorders do not give a
clear picture of the disorders, then the studies
will be flawed.
 Question: How else could we estimate the
prevalence of mental disorders?
13