CBT-I treatment attrition in patients with weekly nightmares.

Abstract

This study's objective was to evaluate the effect of nightmares (NM) on attrition and symptom change following CBT-I treatment using data from a successful CBT-I randomized controlled trial (RCT) delivered to participants with recent interpersonal violence exposure. The study randomized 110 participants (107 women, mean age 35.5), to CBT-I or an attention control group. Participants were assessed at three time periods: Baseline, post-CBT-I (or attention control), and at T3 post-Cognitive Processing Therapy, received by all participants. NM reports were extracted from the Fear of Sleep Inventory. Participants with weekly NM were compared to those with fewer than weekly nightmares on outcomes including attrition, insomnia, PTSD, and depression. Change in NM frequency was examined. Participants with weekly NM (55%) were significantly more likely to be lost to follow up (LTF) post-CBT-I (37%), compared to 15.6% of participants with infrequent NM and were less likely to complete T3 (43%) than patients with less frequent NM (62.5%). NM were unrelated to differential treatment response in insomnia, depression or PTSD. Treatment with CBT-I was not associated with reduced NM frequency; however, change in sleep onset latency (SOL) from post-CBT-I to T3 predicted fewer nightmares at T3. Weekly NM were associated with attrition, but not a reduced change in insomnia symptoms following CBT-I. NM symptoms did not change as a function of CBT-I but change in SOL predicted lower NM frequency. CBT-I trials should screen for NM and consider augmenting CBT-I to specifically address NMs.