Espie, C. A., Fleming, L., Cassidy, J., Samuel, L., Taylor, L. M., White, C. A., . . . & Paul, J. (2008). Randomized controlled clinical effectiveness trial of cognitive behavior therapy compared with treatment as usual for persistent insomnia in patients with cancer. Journal of Clinical Oncology, 26, 4651–4658.
Nurses administered a cognitive-behavioral therapy (CBT) intervention consisting of five weekly 50-minute sessions provided during early afternoon or early evening. The intervention included standard CBT components, such as stimulus control, sleep restriction, and cognitive therapy strategies. The nurses had participated in CBT courses and psychologist-supervised practice and had audiotapes from randomly selected sessions evaluated for congruence with intervention components and principles.
Participants were randomized to either receive CBT or treatment as usual, with stratification for center, prerandomization Pittsburgh Sleep Quality Index (PSQI) scores, existing treatment for insomnia, and tumor type. A 2:1 treatment allocation, in favor of the intervention, was selected to make efficient use of the available CBT sessions and minimize the time patients would have to wait for CBT, thus reducing the potential for patient dropout. Researchers used several strategies to promote intervention fidelity and the integrity of the treatment allocations: the study staff did not interact with other patients in the treatment as usual group, clinicians working with participants in the treatment as usual group did not receive any information about CBT; and printed intervention materials were developed.
This was a randomized, controlled, pragmatic, two-center trial of CBT.
Fatigue Symptom Inventory Interference Subscale
Compared with usual care, CBT resulted in a statistically significant improvement in fatigue interference/daytime fatigue following CBT treatment, and these improvements were sustained at six-month follow-up. Two-thirds of CBT participants attended all therapy sessions, and 94% attended at least three of five CBT sessions. There were similar levels of attrition in the intervention (18%) and usual-care comparison (16%) groups.
Neither interventionists nor patients were blinded to study group allocation, and participants' knowledge that they were assigned to particular treatment arms may have influenced their responses on patient-reported outcome measures.