Understanding the healthcare needs of women Veterans is critical. Insomnia is more common among women than men, with a mean prevalence of 23% among US women, and 54% among women Veterans. Insomnia contributes to lost productivity, psychological distress, medical morbidity, and mortality risk. In prior work we found high rates of treatment discontinuation among women Veterans with insomnia disorder receiving cognitive-behavioral therapy for insomnia (CBT-I). We also found that women Veterans with insomnia preferred non-medication treatments over medications, and that they preferred treatment in individual face-to-face format with a clinical provider. There is a growing literature on treatment of insomnia among individuals with comorbid conditions, suggesting that insomnia treatment may lead to meaningful and durable improvements in sleep quality and other symptoms. Challenges with treatment completion remain, particularly among individuals with significant comorbid conditions, including women Veterans.
This study was a randomized comparative effectiveness trial to evaluate a novel patient-centered insomnia treatment called "Acceptance and the Behavioral Changes to treat Insomnia" (ABC-I) compared to Cognitive-Behavioral Therapy for Insomnia (CBT-I; standard treatment for insomnia). The objectives were: 1) to compare dropout rates and adherence to behavioral recommendations between ABC-I and a similarly-structured CBT-I program, 2) to compare the effectiveness of ABC-I in improving sleep/wake patterns from baseline to post-treatment to a similarly-structured CBT-I program, and 3) to compare the maintenance of improvements 3 months after the end of treatment. A fourth objective was added to develop resources for dissemination of the ABC-I program to VA providers.
759 women Veterans were screened by telephone for eligibility. 347 were enrolled (i.e., provided written informed consent) and completed the screening and baseline assessment visits. 198 women were excluded prior to randomization, most commonly due to severe psychiatric illness or contraindication (n=54; i.e., unstable, bipolar, or active substance use), untreated severe sleep apnea (n=36), or not meeting diagnostic criteria for insomnia disorder (n=34), and 44 individuals withdrew prior to randomization. 149 women were randomized to receive either ABCI (n=75) or CBTI (n=74). The treatments (ABCI or CBTI) were provided in 5 one-on-one sessions by a trained clinical psychologist. Adherence and attrition were measured in both groups. Sleep quality (self-reported and objectively measured), psychiatric symptoms and other measures were assessed at baseline, post-treatment and 3-month follow-up. Differences in adherence to behavioral recommendations between CBTI and ABCI across weeks were assessed via ANOVA with Mixed Effects Models. Differences in treatment completion were assessed using a Fisher's exact test. ANOVA with Mixed Effects Models were used to test differences in sleep improvements (from baseline) between CBT-I and ABC-I groups. Outcomes were: Insomnia Severity Index (ISI), Pittsburgh Sleep Quality Index (PSQI), Epworth sleepiness scale (ESS), and sleep efficiency from actigraphy (objective) and sleep diary (subjective). Analyses were framed to evaluate whether ABC-I was "non-inferior" to CBT-I. In these tests, the null hypothesis (H0:) states that ABC-I is inferior to CBT-I, while the alternative (HA:) states that ABC-I is not inferior to CBT-I. An "inferiority margin" (signified as delta) representing a tolerable margin by which the treatments would not be considered clinically different (i.e., delta) was established a priori for each outcome (e.g., 2-point difference in ISI).
At the conclusion of the trial, we used qualitative methods and expert input to develop a provider training program and to revise patient materials for future dissemination and implementation research.
Aim 1: dropout rates and adherence
There was not a statistically significant difference between the number of participants who dropped out of CBT-I (11%) versus ABC-I (18%; p=.248). Overall dropout rates were somewhat lower than expected in both arms of the trial.
The main adherence outcome was the degree to which participants followed the planned sleep schedule (a key behavioral recommendation in CBT-I and ABC-I). "Adherence" was defined as getting up no more than 15 minutes after the recommended rise time, and going to bed no more than 15 minutes before the recommended bed time. The percent of "adherent" nights were computed for rise and bed times. There were no statistically significant differences between ABC-I and CBT-I adherent rise or bed times during any week of treatment.
Aims 2 and 3: non-inferiority of ABC-I versus CBT-I at post-treatment and 3-months follow-up.
Both treatment groups showed improvements in all key outcome variables from baseline to post-treatment and 3-month follow-up. The main non-inferiority analyses, supported non-inferiority of ABC-I compared to CBT-I. At post-treatment, ABC-I was non-inferior for 3 out of 7 key outcome variables (PSQI sleep efficiency subscale, sleep diary SE, and objective sleep efficiency; p<.016), but non-inferiority was not statistically confirmed for ISI or for PSQI total score, sleep efficiency subscale or daytime dysfunction subscale (p>.063). At 3-month follow-up, ABC-I was non-inferior for all 7 of the key outcome variables (ISI, PSQI total and subscale scores, sleep diary sleep efficiency and objective sleep efficiency; p's<.024).
Aim 4: develop and test ABC-I patient materials and provider training.
Using feedback from three study interventionists, two mental health providers and the ABC-I developer, patient materials, assessments and sleep diaries were modified for clinical use. Based on feedback from mental health providers, the structure of provider education was 5 one-hour web-based live-didactic sessions. Eight providers were invited and 6 participated. Three attended all training sessions, and 3 have provided ABC-I to patients after the training. Two noted that appropriate patients were not referred during the training period, but they anticipate using ABC-I in the future. Two expressed a preference for ABC-I over CBT-I. Provider feedback about the intervention was overall positive.
ABC-I represents a viable alternative to CBT-I for women Veterans, resulting in similar benefits in improved sleep at 3 months. This allows providers to choose the treatment approach that best fits within the clinical context and for patients who do not achieve benefit from CBT-I. Provider perceptions of the treatment support its dissemination.
External Links for this Project
Grant Number: I01HX001287-01A1
None at this time.