Savard, J., Ivers, H., Savard, M.H., & Morin, C.M. (2015). Long-term effects of two formats of cognitive behavioral therapy for insomnia comorbid with breast cancer. Sleep, 39, 813–823. 

To assess the long-term efficacy of a video-based behavioral therapy for insomnia (VCBT-I) as compared to a professionally administered intervention (PCBT-I) and to a no treatment group (CTL).

Participants were randomized to receive one of two types of CBT for insomnia or a control arm with no intervention. The PCBT-I arm received six weekly individual treatment sessions (50 minutes each) with a companion booklet at each session. This was administered by a psychologist or PhD student with CBT experience. The VCBT-I group received a 60-minute video, and they were instructed to watch 5–20 minutes each week and to read a companion booklet each week. They had telephone access to a licensed psychologist if they had questions. All received pre- and postintervention evaluations, as well as long-term follow-up evaluation at 3, 6, and 12 months.

• N = 242  
• AGE = 54.4 years (SD = 8.8 years)
• FEMALES: 100%
• KEY DISEASE CHARACTERISTICS: Study participants had breast cancer within 18 months of completing radiation with insomnia symptoms or using hypnotic medications.  
• OTHER KEY SAMPLE CHARACTERISTICS: 63.9% were married and 92.8% were peri- or postmenopausal.

• SITE: Single site    
• SETTING TYPE: Outpatient  
• LOCATION: Quebec University Hospital, Canada

• PHASE OF CARE: Transition phase after active treatment

• Three-arm randomized, controlled trial

• Insomnia Severity Index (ISI)
• Daily sleep diary: sleep onset latency (SOL), wakefulness after sleep onset (WASO), early morning awakenings (EMA), total wake time (TWT), total sleep time (TST), sleep efficiency (SE) ratio of total sleep time: time spent in bed, and usage of hypnotic medications
• Multidimensional Fatigue Inventory (MFI)
• EORTC Cancer Quality of Life Core 30 (EORTC-QLQ-C30) global quality of life score, item #13
• Dysfunctional Beliefs and Attitudes about Sleep (DBAS)

Post-treatment to follow-up in PCBT-I arm showed significant increase in EMA and TST. The VCBT-I group showed a significant increase in WASA. The control arm showed a significant increase in all sleep variables. There was a reduction, albeit nonsignificant, in the use of hypnotics in the PCBT-I arm, a significant reduction in the VCBT-I arm, and an increase in the control arm. Insomnia remission rates were significantly higher for the PCBT-I group at three and six months when compared to the VCBT-I group. No difference existed at 12 months.

The PCBT-I and VCBT-I groups showed immediate and sustained improvement in several sleep outcomes at three- and six-month follow-ups. When compared, the face-to-face intervention was superior to the video-delivered intervention. The remission rate remained highest in the PCBT-I group. With that said, results should be interpreted with caution because of secondary to selection bias, attrition with significant differences in those who dropped out, and differences in study groups.

• Risk of bias (sample characteristics)
• Selective outcomes reporting
• Findings not generalizable
• Subject withdrawals at 10% or greater
• Not controlling for hypnotic use could affect outcomes/breast cancer only.
• Only 20% of eligible participants agreed to participate, suggesting selection bias.

Further study is needed to document alternative ways to deliver CBT interventions for improving sleep outcomes. Personalized 1:1 CBT appears to be the most effective, although when resources are limited, a video-based intervention can be substituted and does show efficacy.