Using network intervention analysis to explore associations between participant expectations of and difficulties with cognitive behavioural therapy for insomnia and clinical outcome: A proof of principle study

Abstract

Background Research about predictors of response to cognitive behaviour therapy for insomnia (CBT-I) is ongoing. We examined any whether pre-intervention expectations or post-intervention appraisals of difficulties in utilizing face to face (FtF) or digital (dCBT-I) versions of the therapy were associated with outcome. Methods Self-rating data were extracted on 101 adult participants in a recent randomized controlled trial of FtF versus dCBT-I. Network intervention analyses were used to explore any associations between expectations of CBT-I at response at 9 weeks and between post-intervention ratings of difficulties, modality of therapy and response at 9-weeks and at 6-months. Results Anticipated and actual difficulties in employing sleep restriction techniques predicted response in all network models. Modality of therapy played a more overt role in the 9-week outcome network, with FtF therapy more robustly associated with response. However, the direct association between FtF therapy and response was not found in the 6-month outcome network. Notable predictors of poor outcome at 9-weeks and 6-month follow-up were difficulties in accommodating CBT-I into work and daily routines and applying the rules of CBT-I. Conclusions This network intervention analysis highlights that self-confidence and ability in undertaking sleep restriction is a key active ingredient of CBT-I. Also, benefits and gains from access to the FtF version of this multi-component therapy were more apparent in the short than the longer term. However, it is important that findings from this proof of principle study are confirmed in further studies.