Johnson, J.A., Rash, J.A., Campbell, T.S., Savard, J., Gehrman, P.R., Perlis, M., . . . Garland, S.N. (2015). A systematic review and meta-analysis of randomized controlled trials of cognitive behavior therapy for insomnia (CBT-I) in cancer survivors. Sleep Medicine Reviews, 27, 20–28.

DOI Link

Purpose

STUDY PURPOSE: To systematically analyze the available literature and conduct a meta-analysis of randomized, controlled trials (RCTs) that have been completed to date, and to determine a more precise estimate of the efficacy of this intervention on insomnia in people diagnosed with cancer
 
TYPE OF STUDY: Meta-analysis and systematic review

Search Strategy

DATABASES USED: Cochrane CENTRAL, PubMed, PsycINFO, and EMBASE. It is important to note that an ancestry search was performed. Unpublished and ongoing trials were identified through Clinicaltrials.gov and Current Controlled Trials, and authors were contacted to obtain further information.
 
KEYWORDS: Insomnia, cancer, cognitive behavior therapy, randomized controlled trial (combination of MeSH [Medical Subject Headings] and keyword terms)
 
INCLUSION CRITERIA: Study included adults with cancer with clinically relevant insomnia defined by the Diagnostic and Statistical Manual for Mental Disorders, the International Classification of Sleep Disorders, or the Insomnia Severity Index (ISI) with a clinical cutoff score of eight, a selected intervention (CBT-I with both cognitive and behavioral components), and an outcome (sleep diary data reporting on sleep efficiency).
 
EXCLUSION CRITERIA: No specific exclusion criteria was stated.

Literature Evaluated

TOTAL REFERENCES RETRIEVED: The authors screened 1,675 records. Full-text articles were assessed for eligibility (n = 66). Eight studies fulfilled all eligibility for qualitative and quantitative analysis.
 
EVALUATION METHOD AND COMMENTS ON LITERATURE USED: Of the eight studies, five reported on outcomes of women with stage I-III breast cancer, while the remaining reported on the outcomes of men and women with various diagnoses. Interventions were weekly, five to eight weeks in length, and were provided in a wide variety of formats (i.e., individual, group, video, or online). Both passive and active controls were used (i.e., waitlist control, treatment-as-usual, sleep education, behavioral placebo, and mindfulness). A specific scoring system for study quality was used: Quality of methods and design and quality of treatment were independently rated for each RCT. Intra-class correlation coefficient using absolute agreement for two raters was 0.95 for total quality score, 0.94 for treatment quality subscale, and 0.95 for design quality subscale. Meta-analysis was performed on sleep diary and insomnia severity data in seven of eight RCTs.

Sample Characteristics

FINAL NUMBER STUDIES INCLUDED = 8 RCTs
 
TOTAL PATIENTS INCLUDED IN REVIEW: 752 (434 CBT-I and 318 control)
 
SAMPLE RANGE ACROSS STUDIES: 21–150 survivors of cancer
 
KEY SAMPLE CHARACTERISTICS: Five studies reported on outcomes of women with stage I-III breast cancer and the remaining studies reported on outcomes of men and women with various cancer diagnoses.

Phase of Care and Clinical Applications

PHASE OF CARE: Multiple phases of care

Results

CBT-I resulted in a significant improvement on sleep efficiency relative to control from pre- to postintervention (p < 0.01), with a medium effect size pooled across studies (0.53, 95% CI [0.39; 0.68]); and improvement at six-month follow-up (p < 0.01), with effect size (0.33,  95% CI [0.11, 0.54]). There were significant improvement on sleep onset latency relative to control from pre- to postintervention (p < 0.01), with a small to medium effect size pooled across studies (0.43, 95% CI [0.27, 0.58]); and it persisted at six months (p < 0.01), with effect size pooled across studies (0.27, 95% CI [0.11, 0.44]). And, finally, there were significant improvements on wake after sleep onset from pre- to post-intervention (p < 0.01), with a small to medium effect size pooled across studies (0.41, 95% CI [0.24, 0.59]) persisting at six months (p < 0.01), with effect size pooled across studies (0.27, 95% CI [0.11, 0.44]). Overall, there were statistically significant improvements in insomnia severity following CBT-I relative to control from pre- to postintervention (p < 0.01), with a large effect size pooled across studies (0.77, 95% CI [0.6, 0.93]). Four RCTs reported data on insomnia severity at six months (p < 0.01), with effect size pooled across studies (0.54, 95% CI [0.37, 0.73]).

Conclusions

The results of the meta-analysis indicate that survivors of cancer treated with CBT-I showed improvements in sleep efficiency, sleep onset latency, and wake after sleep onset as measured by sleep diaries, and insomnia symptom severity as measured by the ISI when compared to the usual care, waitlist control, or active comparator conditions. In addition, the observed effects persisted at six-month follow-up, suggesting that CBT-I provides significant, lasting improvements in sleep. Future research should examine the mechanisms by which CBT-I improves sleep in the cancer population. And specifically, the relative impact of behavioral versus cognitive change is unclear.

Limitations

It was not possible to directly compare the efficacy of CBT-I as delivered either individually, in a group, online, or via video. Majority of included trials were conducted in women with stage I-III breast cancer, and it is unclear how well the observed results will generalize across the cancer continuum.

Nursing Implications

Individuals diagnosed with cancer are particularly vulnerable to insomnia with a prevalence rate almost twice that of the general population (50%–60% versus 12%–25%). CBT-I was effective across diverse treatment modalities, intervention lengths, and cancer diagnoses or stages in improving sleep diary outcomes that were durable at six months.

Legacy ID

5749