Evaluating the Diagnostic System

463
Classifying Psychological Disorders
psychosocial or environmental problems (such as the loss of a loved one, physical or sexual abuse, discrimination, unemployment, poverty, homelessness, inadequate health care,
or conflict with religious or cultural traditions) that are important for understanding the
person’s psychological problems. Finally, on Axis V, the person is rated (from 100 down
to 1) on current psychological, social, and occupational functioning. Here is a sample
DSM-IV diagnosis for someone who received labels on all five axes:
Axis I: Major depressive disorder, single episode; alcohol abuse.
Axis II: Dependent personality disorder.
Axis III: Alcoholic cirrhosis of the liver.
Axis IV: Problems with primary support group (death of spouse).
Axis V: Global assessment of functioning 50.
Notice that neurosis and psychosis are no longer listed in DSM, because they are not
specific enough. However, some mental health professionals still sometimes use these
terms as shorthand descriptions. Neurosis refers to conditions in which some form of
anxiety is the major characteristic. Psychosis refers to conditions involving more extreme
problems that leave people “out of touch with reality” or unable to function on a daily
basis. The disorders once listed under these headings now appear in various Axis I categories in DSM-IV.
Further changes in the diagnostic system will surely appear in DSM-V, a new edition
of the DSM currently under development for publication in 2010 (Beutler & Malik, 2002;
Helzer & Hudziak, 2002; Maj et al., 2002; McHugh, 2005). For example, some experts
are proposing that diagnoses should consist not just of specific labels, such as “major
depression,” but also of symptom clusters or dimensions that would be rated in terms
of severity (Vollebergh et al., 2001; Widiger & Clark, 2000). The idea behind this dimensional approach would be to create a set of symptom “building blocks” that could be
combined in many different ways so as to better recognize and describe the precise
nature of each person’s problems (Hankin et al., 2005; Krueger & Markon, 2006; Krueger,
Watson, & Barlow, 2005; Markon, Krueger, & Watson, 2005; Widiger & Samuel, 2005).
Evaluating the Diagnostic System
How good is the diagnostic system now in use? One way to evaluate DSM-IV is to consider interrater reliability, the degree to which different mental health professionals agree
on what diagnostic label a particular person should have. Some studies indicate that interrater agreement is as high as 83 percent for schizophrenia and mood disorders; agreement on many other Axis I categories—such as anxiety disorders—is also high (e.g.,
Brown et al., 2001; Nathan & Langenbucher, 1999; Simpson et al., 2002). Still other categories, such as Axis II personality disorders, remain more difficult to diagnose reliably
(Shedler & Westen, 2004; Westen, Shedler, & Bradley, 2006; Zanarini et al., 2000). Overall, interrater agreement appears highest when diagnosis is based on structured interviews
that systematically address each area of functioning and provide guidelines for interpreting people’s responses (Brown et al., 2001; Rogers, 2001; Widiger & Sanderson, 1995).
Do diagnostic labels give accurate information that guides correct inferences about
people? This validity question is difficult to answer because accuracy can be judged in
different ways. A diagnosis could be evaluated, for example, on how well it predicts a
person’s future behavior, or perhaps on whether the person is helped by treatment that
has helped others in the same diagnostic category. Still, evidence does support the validity of most DSM criteria (Clark, Watson, & Reynolds, 1995; Keenan & Wakschlag, 2004;
Kim-Cohen et al., 2005). And validity is likely to improve even more as diagnostic
labels—and the diagnostic system—are refined in DSM-V to reflect what researchers
are learning about the characteristics, causes, courses, and cultural factors involved in
various disorders.
The diagnostic system is far from perfect, however (Beutler & Malik, 2002; Kendell
& Jablensky, 2003; Krueger & Markon, 2006; Nestadt et al., 2005). First, people’s problems often do not fit neatly into a single category. For example, a person may suffer both