Ritterband, L. M., Bailey, E. T., Thorndike, F. P., Lord, H. R., Farrell-Carnahan, L., & Baum, L. D. (2012). Initial evaluation of an internet intervention to improve the sleep of cancer survivors with insomnia. Psycho-Oncology, 21, 695–705.

To determine the feasibility and efficacy of an internet-based, cognitive-based therapy (CBT-I) program—Sleep Healthy Using the Internet (SHUTi)—to improve insomnia symptoms in cancer survivors.

To explore whether an internet-based CBT-I intervention would improve fatigue, mood, and quality of life (QOL).

Patients completed preassessment questionnaires and sleep diaries for 10 days over a two-week period. Patients randomized to the intervention group were enrolled in the SHUTi program. They were given nine weeks to complete six interactive core programs, each lasting about 45 to 60 minutes. During the program, they entered daily sleep diary information. The first core, “Overview,” introduced the SHUTi program, and the next two cores, Behavior 1 and Behavior 2, introduced sleep restriction and stimulus control. The “Education” core focused on sleep hygiene, and the “Sleep Thoughts” core helped users identify and restructure faulty beliefs about sleep. Lastly, patients received “Problem Prevention,” which covered relapse prevention. Upon completion, all patients completed postassessment questionnaires and sleep diaries for 10 days. 

• The sample was comrprised of 28 patients (14% male, 86% female).
• Average age was 56.7 years (standard deviation = 11.7 years).
• Patients were cancer survivors of all types who had, on average, completed treatment about four years prior to the study. Of the patients, 46% had stage I cancer and 21% had stage II cancer.
• On average, the patients had completed 17 years of education, were married (79%), and were white (93%).
• Of the patients in the internet group, 100% were women. In addition, 93% of the internet group had a diagnosis of breast cancer, compared to 36% in control group.

• Patients were recruited from postings at a university cancer center and the surrounding community.
• Home  
• Virginia

• Patients were undergoing the long-term follow-up phase of care.
• The study has clinical applicability for survivorship.

The study was a randomized, controlled trial with a wait-list control.

• Insomnia Severity Index (ISI)
• Sleep diary - self-report
• Multidimensional Fatigue Symptom Inventory–Short Form (MFSI-SF)
• Hospital Anxiety and Depression Scale (HADS)
• Social Funcitioning Health Survey (SF-12)

The internet group showed marked improvement in insomnia severity from pre-/postassessment, and the control group showed no significant change (p < 0.01). The internet group ISI score decreased from 17.1 preassessment to 8.2 postassessment (p < 0.01), and control group scores showed no significant changes from preassessment (15.9) to postassessment (14.4). In addition, the internet group showed clinically significant changes, pre-/postintervention, with 64% having clinically significant insomnia preintervention to 14% having clinically significant insomnia postintervention. Postintervention, 50% of internet group scored “no insomnia” versus 14% in the control group. Effect size for insomnia severity, sleep efficiency, sleep onset latency (SOL), and wake after sleep onset (WASO) were all d > .72.

Improvements were also shown in sleep efficiency and SOL for the internet group. No effect was noted in total sleep time, WASO, and time in bed.

Improvements were seen in the internet group's overall fatigue scores over time (p < 0.01), especially when compared to the control, who showed no improvement over time. There were no significant changes in HADS scores or the SF-12 QOL scores over time or between groups.  
 

An internet-based CBT-I intervention is feasible and efficacious in improving insomnia and fatigue in some cancer survivors in a small sample of highly educated, white, married women. 

• The study had a small, self-selected sample.
• A large percentage of the patients were highly educated, white, married women with breast cancer. The study is not generalizable.

There seemed to be excellent response to this intervention in this population. However, a wider study with a more diverse sample is required.