Naaman, S.C., Radwan, K., Fergusson, D., & Johnson, S. (2009). Status of psychological trials in breast cancer patients: A report of three meta-analyses. Psychiatry, 72, 50–69.

To determine the overall efficacy and magnitude of clinical benefit of psychological interventions in patients with breast cancer, specifically looking at three outcome variables: anxiety, depression, and quality of life (QOL)

Databases searched were MEDLINE (1966–January 2004), EMBASE (1980–2004), Cochrane Controlled Trials Register (1985–February 2004), PsycLit (1973–2004), Biological Abstracts (1990–December 2003), CancerLit (1975–October 2002), CINAHL (1982–December 2003), and Health Star (1975–January 2004).

Search keywords were randomized clinical trial and breast cancer and psychological interventions (cognitive behavioral therapy, group psychotherapy, relaxation, supportive therapy, visual imagery) and psychological adjustment (anxiety, depression, maladjustment, distress, quality of life).

Studies were included in the review if they

• Were a randomized clinical trial (RCT)
• Included at least two arms: an intervention arm and a control arm
• Evaluated the efficacy of a psychological/behavioral intervention
• Were aimed at alleviating psychiatric/psychological morbidity, as defined by anxiety, depression, and/or QOL
• Reported on female patients with a histologically confirmed diagnosis of breast carcinoma of any stage who have undergone surgery.

Trials examining efficacy of interventions designed to assuage surgical distress were excluded.

• A total of 383 citations were identified, with 36 potentially relevant articles identified and screened for retrieval.
• The final meta-analysis included 18 RCTs with usable information by outcome.
• Study quality was evaluated using a framework provided by Cook and Campbell (1979) and a quality assessment scale developed and validated by Jadad and colleagues (1996).

Cook, T.D., & Campbell, D.T. (1979). Quasi-experimentation: Design and analysis issues for field settings. Boston, MA: Houghton Mifflin.

• Fourteen trials assessing anxiety were identified, yielding a sample of 1,278 patients.
• Fourteen trials measuring depression were identified, yielding a sample of 1,324 patients.
• Seven trials measuring QOL were identified, yielding a sample of 623 patients.
• Less than half of the trials included were considered to be of high methodological quality.
• Patients’ ages ranged from 25 to 73 years, and approximately 70% were married or in a committed relationship.
• Studies were conducted in the United States, Canada, England, Italy, Australia, Japan, and China.

Anxiety

• Overall effect size (ES) was -0.40 (95% CI, -0.72 to -0.08) in favor of the treatment condition in comparison to the control.
• Sensitivity analyses exploring the impact of methodological quality on overall ES found a reduction in ES associated with higher quality studies, -0.26 (95% CI, -0.42 to -0.10).
• Trials using patients with high morbidity (metastatic breast cancer) yielded a statistically significant overall ES of -0.40.
• Trials whose treatment extended beyond 20 hours had a statistically significant overall ES of -0.30 in favor of treatment.
• Treatment orientation showed differential impact on overall ES with cognitive behavioral therapy (CBT), yielding an ES of -0.11, as well as ESs of -0.40 for guided imagery and relaxation, -0.43 for supportive-expressive therapy, and 0.02 for educational interventions.

Depression

• Overall ES for depression was -1.01 (95% CI, -1.48 to -0.54) in favor of treatment.
• Trials with higher methodological grade were associated with an overall ES of -0.24 in favor of treatment, while lower quality studies had an overall ES of -1.99.
• Studies in which patients had lower morbidity yielded an overall ES of -0.45, whereas those with patients with more advanced disease yielded an overall ES of -1.20.
• Couples and group therapy reached statistical significance with an ES of -1.02 and -1.35, respectively.
• Treatment orientation showed ESs of -0.85 for CBT, -0.55 for guided imagery, -1.80 for supportive expressive therapy, and -0.45 for educational interventions. With the exception of educational interventions, the other subgroup analyses reached statistical significance.

Quality of Life

• Overall ES was 0.74 (95% CI, 0.12 to -1.37) in favor of the treatment group, but this was not statistically significant. When lower quality studies were removed, statistical significance was achieved, but ES was reduced (0.35, p=0.04).

Overall ES trends among the three outcomes show that more reliable studies were associated with smaller gains. Interventions targeted to patients with clinically important levels of anxiety or depression tended to reap the most benefit, compared to patients who undergo treatment on a prophylactic basis. Group psychotherapy appears to be superior to individual therapy in the treatment of both anxiety and depression. However, a direct impact of group therapy on QOL was not supported in this analysis. CBT interventions appeared to be equally as effective as supportive-experiential therapies. Interventions need not span beyond 20 hours to produce statistically significant ES.

The quality of most studies was not high.

Future trials in psychosocial oncology should incorporate methodological features to enhance internal validity. Evaluation of statistically significant findings on psychometric testing may not reflect clinically significant findings and vice versa. This underscores the need for incorporating qualitative analysis in future studies. There is an absence of studies examining the efficacy of short-term interventions on QOL in advanced breast cancer and should be addressed in future research. Short-term, group interventions may provide the best utilization of scarce resources for the most effect; however, they should be targeted to those patients experiencing clinically important levels of distress. Findings point to the need for higher quality research design and reporting in this field.