|
| Author (s) | Objective and hypothesis listed | Clear eligibility criteria/date study conducted | Settings and sample specified | Interventions clearly described such that replication is possible | Outcome measures specified | Sample size/power analysis | How randomization was accomplished/who randomized | Blinding | Drop out reasons specified | Limitations discussed | Funding source/trial registry | Findings |
|
| Andersen et al. [31]. Randomization was not indicated in title. | The intervention would have a direct positive effect on health (psychological distress, immune function, performance status) | Patients with stage II or stage II breast cancer aged 28 to 84. | A cancer center in Ohio. 227 patients participated. | Patients were assigned to an assessment only (113) or intervention plus assessment arm (114). Patients were also randomized to chemotherapy or radiotherapy. Group sessions were led by female clinical psychologists and consisted of 18 weekly sessions followed by 8 monthly groups (for a total of 1 year). | Impact of Events Scale, POMS, immune measures, Karnofsky Performance Status, symptomatology collected at baseline, 4 months, and 12 months. | No power analysis was included. | Randomization details are discussed in a different publication. | A research nurse blinded to study conditions conducted a health interview with each patient. | Reasons for dropping out were not detailed in this publication. An intent-to-treat analysis was done. | The sample was primarily Caucasian. Little discussion is directed to the study limitations. | Trial register is likely as the study was funded by the National Cancer Institute, the National Institute of Mental Health, the American Cancer Society, and the US Army Medical Research Acquisition Grant, among others. | Functional status increased by 7% in the intervention arm and by 1% in the assessment arm. Symptoms increased by 29% in the assessment arm compared to 14% in the intervention arm. For patients with high cancer-related stress at baseline, declines in mood disturbance were greater in the intervention than the assessment arm. |
|
Andersen et al. [38].
The title specified randomization. | No hypothesis was included in this article. | Patients with stage II or stage II breast cancer aged 28 to 84. Specific recruitment dates were not specified. | A cancer center in Ohio. 227 patients. | A clearly specified intervention with a treatment manual to assure treatment fidelity was used. There was an assessment only (113) or intervention plus assessment arm (114). Patients were also randomized to chemotherapy or radiotherapy. Group sessions were led by female clinical psychologists and consisted of 18 weekly sessions followed by 8 monthly groups (for a total of 1 year). | Likert measures of participant satisfaction and group cohesion were used. In addition the POMS, the Perceived Social Support Scale for Family, the Food Habits Questionnaire, the 7 day physical activity log, the Karnofsky Performance Status Scale of physical functioning, and chemotherapy dose intensity were measured. | Secondary analyses were done. A power analysis was not included. | Randomization details were not included in this article. | Within this article blinding is not discussed. | Reasons for drop out were not specified in this article. | The sample was 90% Caucasian, 9% African American, and 1% Hispanic. Efficacy regarding individual interventions was not possible as social support, stress reduction, and improving mood for example were all included in the group sessions. | The same as Andersen et al. [31]. | Clients were satisfied with the group. Reductions in emotional distress, increases in social support, dietary improvement, reduced variability in chemotherapy dose, improved immunity, fewer symptoms and higher functional status occurred with the group sessions. The intervention did not affect exercise. |
|
Andersen et al. [23].
The title specified randomization. | The psychological intervention would after 11 years reduce risk of disease recurrence and improve survival. | Women diagnosed with breast cancer stage IIA and IIB who were surgically treated and awaiting adjuvant therapy. Recruitment dates were specified. | The same as Andersen et al. [31]. 227 patients participated. | The same as Andersen et al. [31]. | Assessments occurred every 6 months in years 2 to 5 then yearly. Outcomes were recurrence-free survival, breast cancer-specific survival. | The trial was powered to detect a doubling of time to an endpoint requiring 27 events in each treatment arm. | Randomization was according to lymph node status, tumor size, hormone receptor status, menopause status, marital status. Wite and Freedman’s minimization method was used to randomize. Who conducted the randomization was not described. | This article did not mention blinding. | Patients who were lost to followup had their recurrence data censored at the time of last contact. An intent-to-treat analysis was used. Intervention patient, whether they participated or not were included in the analysis. | Little time is spent in the article outlining limitations of the study. | The same as Andersen et al. [31]. | After 11 years patients in the intervention arm had a reduced risk of breast cancer reoccurrence and death compared to those in the assessment arm. Median time to recurrence was 2.8 years for the group and 2.2 years for the assessment only arm. Intervention arm clients survived 6.1 years versus 4.8 years in the assessment arm. |
|
Kissane et al. [19].
The title specified randomization. | The study was not introduced with a hypothesis. | Women with advanced breast cancer (stage IV). Recruitment dates were specified. | Seven public hospitals in Melbourne, Australia. 485 women. | Standardized training and supervision was provided to ensure treatment fidelity. 1 year of clearly defined weekly supportive group therapy was offered by psychologists, psychiatrists and social workers or 3 relaxation therapy sessions. | Survival, depression (Monash Interview for Liaison Psychiatry), quality of life (Quality of Life C-30 questionnaire), Impact of Event Scale and Minimental Adjustment to Cancer Scale. | A power analysis was not presented. A secondary analysis was conducted. | Randomization was by independent adaptive biased coin design. | There was no blinding of patients or therapists. | Reasons for drop out and refusal were given and results were analyzed by intent-to-treat. | Only 47% of eligible patients consented. Results were not reported based on differences in training of therapists nor were the numbers in each group or years of experience presented. It was not discussed whether women practice relaxation at home between sessions. | It is unclear whether there was a trial registry. The study was funded by grants from the National Health and Medical Research Council of Australia, the Cancer Council of Victoria, and the Kathleen Cuningham Foundation. | Group therapy did not prolong survival (the primary outcome) but it prevented and minimized depressive disorders, reduced helplessness, trauma symptoms, and improved social functioning. |
|
Spiegel et al. [36].
The title specified randomization. | Treatment subject would live longer than control subjects. | Women with metastatic and/or recurrent breast cancer who were able to speak English. Recruitment dates were specified. | Kaiser and a university medical center referral in 3 San Francisco bay area communities. 125 women. | It would be difficult to replicate this study. 64 women were randomized to the weekly supportive-expressive group/education session offered by a psychiatrist, psychologist, or social worker. 61 women were randomized to the educational materials group including a 1-year health library membership. | Time to death. | A power analysis was included. | Although a computer-assisted randomization based on a biased coin design was used, the project director and a research nurse conducted the randomization. | Individuals conducting the randomization were not blinded. Women were not blinded. | Reasons for drop out and refusal were given. Analysis was by intent-to-treat. | Different professional led the groups at each site. Attendance at group sessions varied from 1 to 12.5 years. Women in the control group participated in more cancer groups outside the study.
There were baseline differences across sites in age, education, hours worked. | Trial register is likely as the study was funded by the National Cancer Institute, the National Institute of Mental Health, the American Cancer Society, and the Mac Arthur foundation the Fetzer Institute. | The treatment and controlled interventions did not affect survival time. A post hoc analysis revealed estrogen receptor negative clients assigned to the group sessions lived longer than control group clients. |
|
