Support Groups and Survival Detailed Scientific Review

Overview

Background

Current beliefs in the value of support groups for cancer patients have developed out of a half-century of research, anecdotes, scientific theories and studies concerning the effect of family, friends, and other support for responses to life-threatening illnesses and other stress.

In 1975, studies were published for the first time showing that a person’s degree of social support was positively associated with both better adjustment and longer survival. Social support has been variously defined to include people and/or institutions that can help the patient mobilize his or her own inner resources, share tasks or provide services¹.

Organized support groups may be led by specially trained therapists or simply by other patients facing a similar disease challenge. Some groups utilize a variety of therapeutic approaches such as relaxation and the development of coping skills while others rely solely on joining together for exchange of information and emotional support. Some groups exist solely through the Internet^2-4. The rationale and principles of support group interventions for patients with cancer have been discussed by several authors^5-10.

An impetus for the utilization of support groups came in 1989 with the publication of a randomized controlled trial in which women with metastatic breast cancer who participated in a support group had twice the survival time of women in the control group. The support group had received standard medical treatment plus weekly group therapy sessions and lessons in self-hypnosis while the control group received only standard medical treatment¹¹.

Subsequent randomized controlled trials have reported inconsistent effects upon survival^12,13, but many have reported positive effects upon quality of life as described elsewhere in this review^14-16.

Proposed Mechanisms of Action

A preponderance of evidence has linked stress with depression of the immune system and subsequent development of tumors in experimental animals. The stress of major life events such as bereavement and divorce in humans has also been linked with higher mortality in general. The relationship between stress and the incidence and mortality of cancer is less clear. 

"One of the most important ‘buffers’ against the harmful effects of the stress of illness is the presence or availability of persons in the patient’s environment with whom the experience can be shared..."
~~Rowland, Developmental Stage and Adaptation: Adult model. Quoted in ¹.

Adverse Effects

Participating in a support group requires an investment in time and energy that may compete with other activities. Some patients may be upset by the beliefs or coping mechanisms of other group members. Some groups may not be led by appropriately trained individuals or be in contact with reliable sources of information.

Potential Benefits

Studies of potential survival benefits are reviewed in the Summary of Research. Studies of quality of life benefits have not been reviewed by MD Anderson; however, they are included as a subset of a published review of multiple psychological therapies.

Summary of Research

Amount and Type of Research

Based on our review of the literature and other sources between January 1, 1960, and December 31, 2001, we first identified 457 unique English language references to "support groups" or "group support" or "self-help" and "cancer" (including neoplasms, anti-neoplastic and other related terms). Of these articles, 365 (80%) were applicable to formally organized support groups for patients with cancer with just one of these articles reporting a study evaluating survival outcomes.

Five survival studies had previously been identified in the bibliography of an article¹⁷, but only one of those was identified through our initial search. Accordingly, a subsequent search was initiated for the key words "psychotherapy" and "group" combined with cancer related terms. This new search identified 819 articles so it was then restricted to articles concerning survival yielding 89 unique articles in English, of which 17 were applicable to formal group support for cancer patients.

The studies identified in the bibliography (n=6) and both sets of search results (n=1 and 17) were then combined to yield 24 articles less seven articles identified from multiple sources resulting in 14 unique articles that reported or discussed group support and survival outcomes for cancer patients.

Two subsequent references^29,30 were identified, one of which was a new study, and these were added on September 20, 2005. We have classified these 16 articles into the following types of information.

For the human related survival articles, we coded the actual unique outcome studies (9) by the following study designs:

*The National Cancer Institute and other research design authorities consider this to be the strongest type of study design, but this type of study is particularly challenging for support studies. (See Definition of Terms for further information concerning study designs.)

Survival Outcomes

Of eight studies of survival outcomes for patients participating in formally organized support groups, six were randomized controlled trials without blinding^11-16, one was a prospective trial with no controls¹⁸ and one was a retrospective cohort compared with historical controls¹⁹. (An editorial by Spiegel¹⁷ notes a total of 10 randomized controlled trials of psychotherapy for patients with cancer^11-16,20-23, but four of these were of individual rather than group support^20-23.)

Of the six randomized controlled trials (RCTs)^11-16, two studies reported significant positive effects upon survival^11,12 and four others reported no effect^13-16. The two RCTs reporting positive results with support groups described different lengths of support group interventions with the study by Spiegel et al. providing support group meetings for one year¹¹ in contrast to the one by Fawzy et al., in which the support group meetings only continued for six weeks¹². Both studies adjusted for a multitude of potentially confounding variables. Four of the six studies were conducted with women with breast cancer.

The four RCTs reporting negative results included one that attempted to replicate both the length and the type of Spiegel’s original support group intervention¹³. Another trial¹⁵ used Spiegel’s program as a basis for a support group, but added other components and conducted the intervention for a much shorter time (35 weeks versus 52 weeks)¹³. The two remaining trials had other and shorter interventions^14,16.

Three of the four negative trials controlled or used multivariate analysis for potential confounding variables^12,13,15. Although the fourth study presented results within subgroups, it did not present any multivariate analysis in which results are examined for each variable while controlling for each of the other variables. This study also did not have any professionally led groups or provision of specific strategies for pain management¹⁶.

The prospective trial with no controls reported significant differences between sub-groups with more involvement in the support program even though it was only a small exploratory study¹⁸. The retrospective follow-up study did not show any significant benefit after correcting for differences in time between diagnosis and entry into the study^19,24.

Brief notes concerning each of these studies are summarized in the Summary Table for Studies of Support Groups. Additional details may be found in the Annotated Bibliography.

Conclusions concerning the effect of support groups upon survival given two positive studies and four negative studies (three with strong interventions and analysis) are that actual survival effects are doubtful for the general population of patients with cancer, but possible in individual situations.

In response to the news of the latest trial showing no effect, an article in the New York Times reported, "Many women said that they would rejoice if participation in support groups extended survival but that that was not why they joined. ‘I go to my doctor to keep me alive,’ said a woman in Washington. ‘I go to my group to keep me going. And I need both of them.’" (NY Times, Dec. 17, 2001).

Quality of Life Outcomes

A systematic review of quality of life benefits from support groups and other psychological or arts-based approaches may be accessed by returning to Mind-Body Approaches: Multiple Psychological Approaches and Quality of Life.

Annotated Bibliography

Human Studies with Survival Outcomes
Randomized Controlled Trials (RCTs)

¹¹Spiegel D, Kraemer HC, Bloom JR, Gottheil E. Effect of psychosocial treatment on survival of patients with metastatic breast cancer.Lancet 1989;2:888-91.

Purpose: Survival
Type of Study: Review of previous quality of life RCT²⁶
Methods: (Breast cancer) Women (n=109) with metastatic breast were referred to the study investigators by their oncologists. Eighty-six completed the first questionnaire and completed a number of psychological tests; 18 refused to participate, five died before contact. Subjects were then randomly assigned to an intervention (n=50) or a control group (n=36). Of those assigned to the control groups, 14 were too weak at initial interview to participate, six died after entry but before the groups began and two others moved away. Of the controls, 12 were lost to follow-up, four were too ill to participate, two died, four refused to participate and two could not be contacted. The two groups were comparable at study entry; some slight differences in initial stage occurred, but initial stage was not a factor in survival from study entry. Three intervention groups met weekly led by a psychiatrist or social worker with a therapist who had breast cancer in remission. They and the controls also received their usual oncologic care and were followed at four-month intervals. Cox proportional hazard analysis, Kaplan-Meier plots and Chi Square tests was used to examine differences in survival after 10 years.
Results: At the end of 10-year follow-up, three patients remained alive and were contacted by phone and death certificates were obtained for the remainder. Survival did not differ significantly during the first year; but survival time overall from study entry to death was significantly longer for the support groups (36.6 months versus 18.9 months; p<0.0001, Cox; p<0.005, log-rank) as was survival from first metastasis (58.4 months versus 43.2 months; p<0.01, Cox; p<0.04, log-rank). Survival time from initial medical visit to death, however, did not differ significantly. Prognostic factors such as initial stage at diagnosis and types of conventional treatments were controlled for in additional analysis, but the survival differences between the intervention and control groups remained.
Notes: In a subsequent publication, the conventional treatment records of the participants in the study were examined in greater detail. Of the original 86 sets of treatment records, 25 could not be recovered due to having been damaged, lost or discarded. The remaining 61 were determined to be an unbiased sample because their demographics and other medical characteristics were similar to those of the patients whose treatment records were no longer available. No statistically significant differences were found regarding chemotherapy and hormone treatments. The control group did have more adrenalectomies, but this procedure was not associated with any difference in survival²⁷.
Another author compared the survival curve of the control patients to that of a SEER population of breast cancer patients in the same area. The survival curve of the control patients was significantly lower than the general population of breast cancer patients after the point in time when it had diverged from the support group in the study. Thus, the author hypothesized that both differences were simply sampling variations from the normal population curve rather than demonstrations of a support group effect²⁸.

¹³Goodwin PA, Leszcz M, Ennis M, Koopmans J, Vincent L, Guthr H, et al. The effect of psychosocial support on survival in metastatic breast cancer.The New England Journal of Medicine 2001 Dec;345(24):1719-26.

Purpose: Survival, and effects upon mood and pain
Type of Study: Randomized controlled trial
Methods: (Breast) This study attempted to replicate the design of the previous study by Spiegel. Of 235 women recruited into the study, 158 were randomly assigned to the intervention support group and 77 to the control group. Both groups were given educational materials about breast cancer and its treatment, relaxation and nutrition. At base-line and at four, eight and 12 months, the women were asked to complete psychosocial questionnaires. Blinding as to treatment was not done, but psychosocial questionnaires were scored by a research assistant who was unaware of the subjects’ treatment assignment. Univarariate and multivariate Cox proportional-hazards modes were used to assess effects upon survival. The support groups at each of the study centers had leaders who utilized a workbook and went to seminars attended by Spiegel and Classen, authors of the first study that had reported a survival advantage for women with breast cancer who attended a support group¹¹.
Results:

• Compliance and group characteristics: Attendance at group-therapy sessions averaged 66.7%. Participation in support groups outside of the study was reported by 8.2% of women in the intervention group and 10.4% of those in the control group. Support and control groups were balanced in terms of tumor and treatment characteristics. Differences in age, menopausal status, axillary nodal involvement, estrogen receptors, progesterone receptors and the use of adjuvant chemotherapy are discussed in the article. Only progesterone status was associated with survival and higher levels favored longer survival in the intervention group.
• Survival: No significant differences occurred between the support and control groups. Kaplan-Meier median survival was 17.9 months in the intervention group and 17.6 months in the control group. A Cox model was developed that adjusted for progesterone receptors, estrogen receptors, time from first metastasis to randomization, age at diagnosis, nodal stage at diagnosis, use or nonuse of adjuvant chemotherapy, study center, marital status and base-line total mood-disturbance score. The model was not able to identify any significant effects of the intervention on survival (hazard ratio, 1.23, 95% confidence interval, 0.88 to 1.72, P-0.22)
• Profile of Mood States: Baseline scores did not differ, but scores during the first year were significantly lower in the intervention group for total mood disturbance (P=0.02), depression-dejection (P=0.002), tension-anxiety (P=0.002), anger-hostility (P=0.007) and confusion-bewilderment (P=0.02). Adjustment for baseline inequalities diminished these differences, but revealed a significant interaction between base-line score and effect of intervention. Women who were initially more distressed benefited from the support group intervention (13.5% of scale range), but women who were not as distressed did not benefit.
• Experience of Pain and Suffering or Hurt: On two visual analogue scales, women in the intervention group reported less worsening of pain over the course of one year than controls (15% of scale range, P=0.04). A significant interaction with baseline scores occurred in which women in the intervention group benefited only if their baseline scores were high. No other significant differences occurred in the experience of suffering or hurt.

¹²Fawzy FI, Fawzi NW, Hyun CS, Elashoff R, Guthrie D, Fahey JL, et al. Malignant melanoma: Effects of an early structured psychiatric intervention, coping, and affective state on recurrence and survival 6 years later. Archives of General Psychiatry 1993 Sep;50(9):681-89.

Purpose: Survival
Type of Study: Randomized controlled trial
Methods: (Skin, melanoma) All patients referred for treatment of malignant melanoma were screened for study eligibility such as stages I, no metastasis, or II, local nodal disease (other criteria detailed in article). Eighty patients agreed to participate and half of them were randomized to the support group intervention. By the end of five years after various dropouts, 34 support group patients and 34 controls remained. Treatment for all patients was limited to surgical removal; no chemotherapy or radiation was given. Patients in the control group did not receive any psychiatric intervention or have contact with any of the support group leaders. The support group program consisted of a six-week structured psychiatric group intervention with small groups (seven to 10) meeting weekly. Support consisted of education concerning melanoma and basic nutrition, stress management, enhancement of coping skills and psychological support from group and staff. Attendance at support meetings was "almost 100%". Time to recurrence and survival was tested with Kaplan-Meier and Cox proportionate hazards regression models.
Results: After five years, a larger proportion of the control patients had recurrences (38% versus 21%, P<0.09 for trend log rank test) and a smaller proportion of the controls had survived (71% versus 91%, P=0.03 log rank test). The survival of the control patients was found to be similar to that of matched controls from the hospital’s database of 6000 melanoma patients which verified the reliability of these controls as a representative sample. Covariates (older age, male sex, Breslow depth greater than 1.5 mm and tumor site on trunk) were entered along with support group intervention into a Cox regression model. Only Breslow depth and support group remained as significant factors for recurrence and survival with support group participation remaining significant after adjusting for Breslow depth (P= 0.0438 for recurrence and 0.0066 for survival). Additional development of a parametric model also indicated that Breslow depth and support group were both significant factors for recurrence and survival.
Notes: A subsequent article has been published with results from a 10-year follow-up²⁹. At 10 years, participation in the intervention appeared to have no effect on recurrence. It still had a significant effect for survival (p=.05); however, the survival benefit of the support group had weakened since the five-year follow-up. As noted by the authors, the study was originally designed only to assess the effects of the intervention on enhanced coping and decreased distress. This difference in original intent and the small sample size, limits the validity of any generalizations about recurrence and survival that can be based on this study.
A replication of this study was conducted by  Boesen et al.³⁰; however, only distress and coping capacity were evaluated, not survival or recurrence. Patients in the intervention group showed significantly less fatigue (p=.04), greater vigor (p=.003) and lower total mood disturbance (p=.001), and used significantly more behavioral and cognitive coping strategies than the control group (p=.0007 and p=.0002, respectively). Improvements were only significant at the first follow-up of six months.

³⁰Boesen E, Ross L, Frederiksen K, Thomsen B, Dahlstrom K, Schmidt G, et al. Psychoeducational intervention for patients with cutaneous malignant melanoma: a replication study. Journal of Clinical Oncology 2005 Feb;23(6):1270-7.

Purpose: Evaluate the effects on distress and coping capacity
Type of Study: Replication study of a RCT¹²
Methods: (Skin, melanoma) Patients 18 to 70 years of age with cutaneous malignant melanoma were contacted in an outpatient clinic in Denmark three to 12 weeks after surgery. Out of 420 eligible patients, 399 met the inclusion criteria, and 262 consented (66%). Reasons given for refusal were the distance involved, lack of time and no need for support. Out of the 131 patients randomly assigned to the intervention group, 19 were excluded and/or dropped out due to treatment, death, time and distance for involvement and not needing support. Of the 131 patients randomly assigned to the control group, two dropped out due to disappointment over control group status and an unknown reason. The intervention group (n=112) grogram consisted of six 2 ½-hour group sessions over a period of six weeks based on the manual developed by Fawzy et al.¹² Changes in the intervention program from the original study included the health education component being provided by specialized health staff instead of the group therapist, and teaching coping strategies through discussions of the patients’ own experiences rather than using coping scenarios. Patients in the control group did not receive any psychoeducational intervention. All participants completed a questionnaire at baseline and six and 12 months after surgery. For patients in the intervention group, time between the end of the intervention and the first follow-up questionnaire ranged from 1 to 60 days.
Results: No significant differences were noted between groups at baseline. Comparing participants and dropouts, more women dropped out (p=.02). Patients in the intervention group showed significantly less fatigue (p=.04), greater vigor (p=.003), and lower total mood disturbance (p=.001) compared with controls at six months. The intervention group also used significantly more behavioral and cognitive coping strategies than the control (p=.0007 and p=.0002, respectively) at six months.

¹⁴Edelman S, Lemon J, Bell DR, Kidman AD. Effects of Group CBT on the survival time of patients with metastatic breast cancer. Psycho-Oncology 1999;8:474-81. (One of two reports concerning one study.)

Purpose: Survival
Type of Study: Randomized controlled trial
Methods: (Breast cancer) Of 124 patients with metastatic breast cancer, 62 were randomly assigned to a support group intervention program and 62 to be controls. Subjects were recruited over a three-year period (March 1994 to February 1997) and survival analysis was conducted in 1999; thus patients were followed between two and five years. Two patients in the support group and one in the control group were later found to not have metastatic disease and were excluded from the study. Sixteen patients died before or during the study period, 10 became too ill to attend and three had other problems that caused them to be dropped from the study, yielding 92 evaluable patients. Support group sessions were held for eight weeks, followed by a family night and then three more monthly meetings. Survival time was calculated from the entry to the study using Cox Proportional Hazards regression to identify factors associated with survival time. Cox model development began with support group membership plus seven other factors that had been previously shown to be associated with survival: stage at initial diagnosis, lymph node status at diagnosis, ECOG performance status at trial entry, disease status at trial entry, visceral or non-visceral metastases at trial entry, chemotherapy within two weeks of trial entry and use of hormonal medication at trial entry. (Radiotherapy was not included because of the small number of patients undergoing it.)
Results:

• Demographics and Disease Characteristics: No significant differences (all P values >0.1).
• Survival: No significant differences in Kaplan-Meier survival curves from time of recruitment or from time of diagnosis of metastasis. Support group membership was not significantly related to survival time from either recruitment or date of recurrence according to Cox regression analysis. The best fitting Cox model of survival from recruitment revealed that three indicators related to survival:
  • Presence or absence of visceral metastasis
  • ECOG performance score at trial entry
  • Chemotherapy at trial entry
• Psychological Outcomes: Psychological outcomes for this study were reported in a separate article²⁵.

Note: This study differed from the earlier study by Spiegel¹¹ in the shorter length of the intervention and the emphasis within the support group program upon acquisition of cognitive and behavioral coping skills rather than the expression of thoughts and feelings.

¹⁵Cunningham AJ, Edmonds CVI, Jenkins GP, Pollack H, Lockwood GA, Warr D. A randomized controlled trial of the effects of group psychological therapy on survival in women with metastatic breast cancer. Psycho-Oncology 1998;7:508-17.

Purpose: Survival
Type of Study: Randomized controlled trial
Methods: (Breast cancer) Women with metastatic breast cancer (n=66) of 246 eligible patients with metastatic (other than brain) breast cancer agreed to participated in the study. Thirty patients were randomly assigned to the intervention group and 36 to the controls. Randomization was stratified by age (50 and older or younger) and visceral versus non-visceral metastases. The group psychological intervention consisted of 35 weekly two-hour sessions of supportive plus cognitive behavioral therapy. The principles of the support group program were those of the original support group study as described by Spiegel et al. but they further developed and standardized their intervention through four initial pilot support groups. Each of the small support groups was led by a trained and experienced psychotherapist. The control group received a home study cognitive behavioral package including a workbook and two audiotapes for self-directed practice of relaxation and mental imaging with phone calls from the study coordinator every two months for a year. Both groups received "standard medical care". Subjects entered into the study during a two year period (1992-1994), and follow-up continued until five years after the first subjects were entered. Mann-Whitney tests were used to test whether the groups were balanced in terms of time from primary disease to diagnosis of metastatic disease, from diagnosis of metastatic disease to randomization and from diagnosis of primary disease to randomization. Log rank tests compared Kaplan-Meier survival curves from date of randomization, first metastatic diagnosis and first diagnosis of breast cancer to the end of the fifth year of the study.
Results: The groups were balanced in terms of demographics other than education with the intervention group averaging more education. The support group was also more likely to listen to audiotapes for relaxation (p<0.05) and have specific dietary habits related to health (p<0.10). in the primary log rank test, no significant differences were found in survival between the support and intervention groups. A slight trend in favor of the intervention group disappeared when controlling for the length of time from the first diagnosis of metastatic disease (support group median survival 44.8 months compared with 43.8 months for the controls, P=0.79). The only variables found to be significantly predictive of survival were exercise (P=0.014) and attendance at an outside support group by the sub-group of subjects within the study support group (P=0.05).
Note: Further analysis of survival is planned 10 years from the start of the study.

¹⁶Ilnyckyj A, Farber J, Cheang MC, Weinerman BH. A randomized controlled trial of psychotherapeutic intervention in cancer patients. Ann R Coll Physicians Surg Can 1994;27:93-6

Purpose: Survival
Type of Study: Randomized controlled trial
Methods: (Various cancers) Patients with any type of malignancy were invited to participate in the trial and 401 patients were interviewed, of which 127 signed consent forms and five were excluded for various reasons. Half of the patients had breast cancer; the remaining cancer types are not noted. Randomization to one of three different support groups or control status was stratified by gender, performance status and disease status. One of the three support groups was led by a social worker for six months (n=31), another group was led by a social worker for three months and then met for three months without (n=30) and a third group met for six months without any leadership at all (n=35). The 26 (?) control patients did not participate in any groups. Group leaders were not instructed in any specific techniques other than to give information and "be supportive". A group of 21 patients was added to the group without leadership for only the last three months of the study to replace the unexpectedly high attrition rate from this group. Analysis was conducted with and without these later additions. For survival analysis, the two groups with any leadership were combined into one group. Psychological tests given at pre-randomization and at post-randomization intervals consisted of state trait anxiety inventory, Beck depression scale and the health locus of control scale. No statistically significant differences existed in prognosis (as judged by an oncologist) or in demographics.
Results: A survival difference could not be shown for the support groups either with or without leadership. Kaplan-Meier curves showed no differences. Mean survival was 70.7 months for the leadership groups and 62 months for the group with no leadership compared with 82.4 months for controls (P=0.53, log-rank).
Notes: Treatment with surgery, radiation or chemotherapy is not described in the article.
Although Cox proportional hazard modeling was reportedly used, results of this analysis are not reported in the article.
Results of the various psychological tests are not described in this article.

Non-randomized Survival Studies

¹⁸Cunningham AJ, Phillips C, Lockwood GA, Hedley DW, Edmonds CV. Association of involvement in psychological self-regulation with longer survival in patients with metastatic cancer: an exploratory study. Advances in Mind-Body Medicine 2000 Fall;16(4):276-87.
(Article not obtained; review based upon abstract)

Purpose: Survival
Type of Study: Prospective with no controls (exploratory study)
Methods: (Various cancers) Participants were 22 patients with various kinds of metastatic cancers who participated in one year of weekly group psychological therapy. Verbal data from patients’ written homework and therapist notes were collected and qualitatively classified for low, moderate or high involvement in the program. Predicted survivals did not differ significantly between the groups as judged by a panel of oncologists.
Results: A significant beneficial relationship was found between the subgroups with more involvement in and survival duration as evidenced by Kaplan-Meier analysis (P=0.006, Log Rank test). Likely confounders of medical status, age, quality of life, and attendance at therapy were similar across the subgroups and did not change these results.

¹⁹Gellert GA, Maxwell RM, Siegel BS. Survival of breast cancer patients receiving adjunctive psychosocial support therapy: a 10-year follow-up study. J of Clinical Oncology 1993;11:66-9.
(Abstract only)

Purpose: Survival
Type of Study: Retrospective follow-up study
Methods: (Breast cancer) Participants (n=34) were retrospectively matched to 102 non-participants on several prognostic factors and followed until the end of the study. The support program consisted of weekly cancer peer support, family therapy, individual counseling and use of positive mental imagery. They were then monitored from the dates of diagnosis (1971 through 1980) until March of 1991. According to the authors, survival analysis controlled for major prognostic factors.
Results: Mean survival time from diagnosis was 96 months (+/- 53.2 months) for participants versus 85.1 months (+/- 63.4 months) in the controls, a non-significant difference (P=.1). Median survival difference was also not significant with 84.0 months for participants (95% C.I., 59-133) and 66.0 months for controls (95% CI, 48 – 105).
Comment: These are extremely wide standard deviations and confidence intervals that indicate an extreme variance in survival times within both the participant and the control groups. This could indicate many different factors influencing survival that may or may not have been included in the study and/or analyzed with appropriate multivariate analysis techniques not described in this abstract.
Note: A previously published article²⁴ by Morgenstern and colleagues had reported results for this same cohort 10 years earlier. At that time, preliminary findings suggested a strong beneficial effect of the support program on survival. However, when correction was made for the differing lag times between date of diagnosis and the start of the program, the benefit was no longer significant. Accordingly, correction for these lag times was presumably included in the next 10-year follow-up study.
Caution Concerning This Report: This report was not published in a peer-reviewed journal; that is, it was not reviewed by experts in the field to determine that there was sufficient data to support the conclusions.

Reference List

Effects of Support Groups upon Survival

1. Lerner M. Psychological Approaches to Cancer. Lerner M. Choices in Healing: Integrating the best of conventional and complementary approaches to cancer. 1st ed. Cambridge, MA: The MIT Press, 1994:137-94.
2. Klemm P, Hurst M, Dearholt SL, Trone SR. Gender differences on Internet cancer support groups. Computers in Nursing 1999 Mar-1999 Apr;17(2):65-72.
3. Klemm P, Reppert K, Visich L. A non-traditional cancer support group. The Internet. Computers in Nursing 1998 Jan-1998 Feb;16(1):31-6.
4. Sharf BF. Communicating breast cancer on-line: support and empowerment on the Internet. Women & Health 1997;26(1):65-84.
5. Holland JC. Psycho-oncology: Overview, obstacles and opportunities. Psycho-Oncology 1992;(1):1-13.
6. Leszcz M, Goodwin PJ. The rationale and foundations of group psychotherapy for women with metastatic breast cancer. International J of Group Psychotherapy 1998;48(2):245-73.
7. Krupnick JL, Rowland JH, Goldberg RL, Daniel UV. Professionally-led support groups for cancer patients: an intervention in search of a model. International Journal of Psychiatry in Medicine 1993;23(3):275-94.
8. Bottomley A. Psychotherapy groups and cancer patient survival: chasing fool's gold? European Journal of Cancer Care (English Language Edition) 1998;7(3):192-6.
9. Mathews HF. Negotiating cultural consensus in a breast cancer self-help group. Medical Anthropology Quarterly 2000;14(3):394-413.
10. Fawzy FI. A short-term psychoeducational intervention for patients newly diagnosed with cancer. Support Care Cancer 1995;3:235-8.
11. Spiegel D, Kraemer HC, Bloom JR, Gottheil E. Effect of psychosocial treatment on survival of patients with metastatic breast cancer. Lancet 1989;2:888-91.
12. Fawzi FI, Fawzi NW, Hyun CS, Elashoff R, Guthrie D, Fahey JL, et al. Malignant melanoma: effects of an early structured psychiatric intervention, coping, and affective state on recurrence and survival 6 years later. Archives of General Psychiatry 1993 Sep;50(9):681-89.
13. Goodwin PA, Leszcz M, Ennis M, Koopmans J, Vincent L, Guthr H, et al. The effect of group psychosocial support on survival in metastatic breast cancer. The New England Journal of Medicine 2001 Dec;345(24):1719-26.
14. Edelman S, Lemon J, Bell DR, Kidman AD. Effects of group cbt on the survival time of patients with metastatic breast cancer. Psycho-Oncology 1999;8:474-81.
15. Cunningham AJ, Edmonds CVI, Jenkins GP, Pollack H, Lockwood GA, Warr D. A randomized controlled trial of the effects of group psychological therapy on survival in women with metastatic breast cancer. Psycho-Oncology 1998;7:508-17.
16. Ilnyckyj A, Farber J, Cheang MC, Weinerman BH. A randomized controlled trial of psychotherapeutic intervention in cancer patients. Ann R Coll Physicians Surg Can 1994;27:93-6.
17. Spiegel D. Mind Matters: Group therapy and Survival in Breast Cancer. New England Journal of Medicine 2001; 345(December):1767-8 .
18. Cunningham AJ, Phillips C, Lockwood GA, Hedley DW, Edmonds CV. Association of involvement in psychological self-regulation with longer survival in patients with metastatic cancer: an exploratory study. Advances in Mind-Body Medicine 2000 Fall;16(4):276-87.
19. Gellert GA, Maxwell RM, Siegel BS. Survival of breast cancer patients receiving adjunctive psychosocial support therapy: a 10-year follow-up study. J of Clinical Oncology 1993;11:66-9.
20. Richardson JL, Shelton DR, Krailo M, Levine AM. The effect of compliance with treatment on survival among patients with hematologic malignancies. Journal of Clinical Oncology 1990 Feb;8(2):356-64.
21. Kuchler T, Henne-Bruns D, Rappat S, Graul J, Holst K, Williams JI, et al. Impact of psychotherapeutic support on gastrointestinal cancer patients undergoing surgery: survival results of a trial. Hepato-Gastroenterology 1999;46:322-35.
22. Ratcliffe MA, Dawson AA, Walker LG. Eysenck personality inventory l-scores in patients with hodgkin's disease and non-hodgkin's lyphoma. Psycho-Oncology 1995;4:39-45.
23. Linn MW, Linn BS, Harris R. Effects of counseling for late stage cancer patients. Cancer 1982;49:1048-55.
24. Morgenstern H, Gellert GA, Walter SD, Ostfeld AM, Siegel BS. The impact of a psychosocial support program on survival with breast cancer: The importance of selection bias in program evaluation. Journal of Chronic Disease 1984;37:273-82.
25. Edelman S, Bell DR, Kidman AD. A group cognitive behaviour therapy programme with metastatic breast cancer patients. Psycho-Oncology 1999;1999(8):295-305.
26. Spiegel D, Bloom JR, Yalom I. Group support for patients with metastatic cancer. Arch Gen Psychiatry 1981 May;38:527-33.
27. Kogon MM, Biswas A, Pearl D, Carlson RW, Spiegel D. Effects of medical and psyhotherapeutic treatment on the survival of women with metastatic breast carcinoma. Cancer 1997 Jul;80(2):225-30.
28. Fox BH. A hypothesis about Spiegel et al.'s 1989 paper on psychosocial intervention and breast cancer survival. Psycho-Oncology 1998;7:361-70.
29. Fawzy F, Canada A, Fawzy N. Malignant melanoma: effects of a brief , structured psychiatric intervention on survival and recurrence at 10-year follwo-up. Arch Gen Psychiatry 2003 Jan;60(1):100-3.
30. Boesen E, Ross L, Frederiksen K, Thomsen B, Dahlstrom K, Schmidt G, et al. Psychoeducational intervention for patients with cutaneous malignant melanoma: a replication study. Journal of Clinical Oncology 2005 Feb;23(6):1270-7.