DSM-5's Validity: Non Sumus Angeli!

Ronald W. Pies, MD


June 12, 2013

In This Article

More Than a Lab Test

It is probably too early to know how the modifications in the DSM-5 will affect these validity measures compared with DSM-III and DSM-IV criteria. And, it could be argued that more time should have been taken to find out. Nonetheless, historically, several of the major diagnostic categories in the DSMs have shown both good reliability (kappa values, or interrater agreement) and reasonably good predictive validity. For example, Mason and colleagues[4] found that both DSM-III-R and ICD-10 diagnoses of schizophrenia had high predictive validity for long-term outcome, and that both provided relatively stable diagnoses over time.

Similarly, in an 8-year, prospective study of bipolar I disorder in children,[5] there was strong evidence of stability and continuity of the diagnosis, into adulthood. For example, in adult individuals with childhood bipolar I disorder, the 44.4% frequency of manic episodes was 13-44 times higher than in the general population -- an indicator of good predictive validity of the bipolar I diagnosis.*

Dr. Carroll believes that many other psychiatric disorders have achieved good convergent validity, including mania, melancholia, panic disorder, obsessive-compulsive disorder, and catatonia.[3] Nonetheless, the next update of the DSM-5 needs to include more convergent validity data for its diagnostic categories, including how well they predict course of illness, morbidity and mortality, comorbid illness, and response to treatment.

We can do better than the DSM-5, and we must strive for ever-increasing levels of validity -- but this is much more than a matter of fixating on laboratory tests. Psychiatry is, and will remain, a clinical science, unlike physics or thermodynamics. And regardless of whether laboratory are available, there is no substitute for listening carefully to the patient's narrative, getting a comprehensive family history, ruling out pertinent medical and neurologic confounders, and using sound clinical judgment in making the diagnosis.

The bottom line belongs to Dr. Carroll: "Laboratory measures are the servants of clinical science, not the other way around."[3]

Acknowledgments. The author thanks Dr. Bernard Carroll and Dr. Allen Frances for their comments on early drafts of this commentary; however, the author alone is responsible for the conclusions herein.

*The Geller study used a more conservative criterion for manic episodes than the DSM-IV does: 2 weeks of manic symptoms were required, rather than 1, along with at least 1 "cardinal symptom" of mania (elation and/or grandiosity). Furthermore, the study's rigorous evaluation of the children and their parents is hardly representative of everyday outpatient practice, where increasing prescription of antipsychotic medication in children and adolescents is a serious concern.[6] Nevertheless, the Geller study is a good example of predictive validity in psychiatric diagnosis.

Suggested Reading

Carroll BJ. Diagnostic validity and laboratory studies: rules of the game. In: Robins LN, Barrett JE, eds. Validity of Psychiatric Diagnosis (American Psychopathological Association//Proceedings of the Annual Meeting). New York: Raven Press; 1989.

Lux V, Kendler KS. Deconstructing major depression: a validation study of the DSM-IV symptomatic criteria. Psychol Med. 2010;40:1679-1690.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: