Publication Date




Embargo Period


Degree Type


Degree Name

Master of Science (MS)


Psychology (Arts and Sciences)

Date of Defense


First Committee Member

Annette M. La Greca

Second Committee Member

Gail H. Ironson

Third Committee Member

Susan K. Dandes


Posttraumatic stress disorder (PTSD) is a frequently studied outcome in children after natural disasters. However, different criteria for diagnosing PTSD have been recommended by the 5th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) and the proposed 11th edition of the International Classification of Diseases (ICD-11). DSM-5 advocates a broad approach to PTSD diagnosis, whereas ICD-11 proposes a narrow definition of PTSD. Although children are particularly vulnerable to PTSD following disasters, little is known about whether these revised criteria are appropriate for children, as diagnostic revisions were based mostly on adult research. The present study investigated rates of PTSD using DSM-IV, DSM-5, and ICD-11 criteria in two samples of children (ages 7 – 11) affected by a natural disaster. Children exposed to Hurricane Ike (n = 327) or Hurricane Charley (n = 383) completed self-report measures eight to nine months postdisaster. Diagnostic algorithms for DSM-IV, DSM-5, and ICD-11 (two versions) were used to calculate rates of “probable” PTSD. ICD-11 (Version 1) identified the most children with “probable” PTSD for Ike (18.3%) and Charley (10.7%). The omission of intrusive memories in ICD-11 criteria caused the Re-experiencing cluster to preclude the most children from meeting diagnosis, which differs markedly from DSM-IV and DSM-5 patterns of cluster endorsement. DSM-5 identified the fewest children for Ike (14.1%) and Charley (6.5%), indicating that a four-factor model of PTSD may be inappropriate for children. The DSM-5 Cognitions/Mood cluster was met by the fewest children, suggesting that these symptoms may not be developmentally sensitive. Of the children identified by any diagnostic system as having PTSD, only about a third were identified by all systems, indicating low agreement among systems. Children identified by all systems could be differentiated by age, perceived life threat, and actual life threat. Several risk factors identified by prior research (e.g., race/ethnicity, actual life threat, and immediate loss/disruption) were predictive of PTSD across diagnostic systems, although perceived life threat and negative life events were predictive of ICD-11 but not DSM-IV or DSM-5. These findings provide support for the ICD-11 proposal and question the suitability of DSM-5 criteria for preadolescent children, although additional research on developmentally-appropriate criteria is needed.


PTSD; Natural Disaster; DSM-IV; DSM-5; ICD-11