0499 Predictors of Session Attendance in a RCT for CBT-I for Perinatal Insomnia

Abstract

Abstract Introduction Cognitive behavioral therapy for insomnia (CBT-I) is recommended as the first-line treatment for chronic insomnia disorder. However, early treatment dropout can negatively impact the success of treatment. In order to design effective strategies to reduce attrition we need to understand predictors of dropout. In this work, we focus on predictors of treatment session attendance among pregnant women in a randomized controlled trial of CBT-I. Methods Participants were 87 pregnant women with insomnia disorder (mean age=32.5, SD=5.1 years) who were enrolled in a randomized controlled trial of CBT-I and were randomized to the CBT-I arm (5 sessions). We did not include women who did not complete treatment due to early labor. The Insomnia Severity Index (ISI) and Edinburgh Postnatal Depression Scale (EPDS) and demographic questionnaires (age, income, and educational background) were administered at screening. Results Logistic regression analyses were conducted to identify predictors of number of CBT-I sessions attended during pregnancy. A logistic regression model that included clinical predictors found that the ISI and EPDS were not significant predictors of session attendance. A logistic regression model that included demographic predictors (income, education, and age) was significant (F(3,76)=6.49, p<0.001) with an R² of .204. Independently, income was not a significant predictor (β =.15, p=.32), but education (β =-.21, p<.05) and age (β =-.48, p<.01) were significant predictors of fewer sessions attended. Dropping income from the model, there was a significant age by education effect (β =1.39, p<.05). Among participants with less than a college degree, those who were younger had attended fewer sessions during pregnancy. Among those who completed a college education and above, number of sessions attended did not differ by age. Conclusion Pregnant women with insomnia who were randomized to receive CBT-I were more likely to withdraw early from treatment if they were younger and had less than a college education. Further research that focuses on increasing attendance in CBT-I treatment among pregnant women needs to develop strategies to increase retention in this vulnerable population. Support NR013662