Citation

Schwarz ED, Kowalski JM. Am. J. Psychiatry 1991; 148(5): 592-597.

Affiliation

Division of Child and Adolescent Psychiatry, Evanston Hospital, Ill.

Copyright

(Copyright © 1991, American Psychiatric Association)

DOI

unavailable

PMID

2018159

Abstract

OBJECTIVE: The purpose of the study was to investigate the effect of symptom threshold and criteria set selections on the diagnosis of posttraumatic stress disorder (PTSD) in adults and children exposed to a man-made disaster and determine how well DSM-III and its successors agree. METHOD: Data gathered in the course of a voluntary clinical screening for PTSD in 66 adults and 64 children 6 to 14 months after exposure to a school shooting were analyzed according to DSM-III, DSM-III-R, and proposed DSM-IV criteria for PTSD diagnosis and cluster endorsement using liberal (occurring at least a little of the time), moderate (occurring at least some of the time), and conservative (occurring at least much or most of the time) symptom thresholds. RESULTS: Within DSM-III, DSM-III-R, and proposed DSM-IV, selection of liberal, moderate, and conservative symptom thresholds had robust effects on rates of diagnoses; liberal thresholds allowed the greatest frequencies of diagnosis. Compared with DSM-III and proposed DSM-IV, DSM-III-R generally diagnosed the fewest cases. Agreements between DSM-III-R and proposed DSM-IV were good, while agreements between DSM-III and its successors varied for children and adults. CONCLUSIONS: Diagnostic rates and agreements were complexly influenced by interactions among threshold and revisions in symptom clusters. The present study suggests that attempts to refine PTSD classification consider specification of symptom threshold intensity and supports the view that modification of criteria sets be undertaken with caution.

Language: en