al., 1995, 199719). The committee also identified a progress report for an ongoing 3-year study with selected results; only preliminary findings were available so it was not included on the review (Boudewyns and Hyer, 1996). See Tables 4-4 and 4-5 for a summary of included studies.


The committee identified three RCTs of cognitive restructuring compared with coping skills training or an educational booklet. One study suffered high dropout rates (up to 46 percent) and used LOCF to address missing data. Another study with no major limitations showed no difference between cognitive restructuring and exposure-focused therapy, but had no control group (Tarrier et al., 1999). A second trial with no major limitations conducted in individuals who had experienced a motor vehicle accident had a modest dropout rate handled by LOCF, and showed significant improvement on CAPS and loss of diagnosis (Ehlers and Clark, 2003). However, the committee was reluctant to judge cognitive restructuring on the basis of this single trial in victims of motor vehicle accidents.

Synthesis: The committee judged the overall body of evidence on cognitive restructuring in the treatment of PTSD to be moderate quality, but there were important limitations. Although the three studies identified were all of medium size and two were reasonably well-conducted, one of the two did not find an effect and the other found a large effect. The committee is uncertain about the presence of an effect, and believes that future well-designed studies will have an important impact on confidence in the effect and the size of the effect.

Conclusion: The committee concludes that the evidence is inadequate to determine the efficacy of cognitive restructuring in the treatment of PTSD.

Exclusion Notes

The committee did not identify any studies on cognitive restructuring alone to exclude. See Table 4-6 for a summary of included studies.


Less than half of the 2005 study participants met PTSD diagnosis. Separate results for those with and without PTSD not provided except for one supplemental analyses for those with PTSD. The 1997 follow-up study did analyze PTSD versus non-PTSD patients separately, however there was no longer a control group so the study is uninformative with regard to the core question of efficacy.

The National Academies of Sciences, Engineering, and Medicine
500 Fifth St. N.W. | Washington, D.C. 20001

Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement