Annals of Surgical Oncology

, Volume 19, Issue 2, pp 373–378

Improving Symptoms and Quality of Life of Female Cancer Survivors: a Randomized Controlled Study


    • Marilyn and Walter J. Wolpin Comprehensive Breast Care CenterBeaumont Health System
    • The OU William Beaumont School of MedicineOakland University
  • Robert Jarski
    • The OU William Beaumont School of MedicineOakland University
    • School of Health SciencesOakland University
  • Heather Rea
    • School of Health SciencesOakland University
  • Ronald Gellish
    • School of Health SciencesOakland University
  • Frank Vicini
    • The OU William Beaumont School of MedicineOakland University
    • Oncology ServicesBeaumont Health System
Healthcare Policy and Outcomes

DOI: 10.1245/s10434-011-2051-2

Cite this article as:
Lerman, R., Jarski, R., Rea, H. et al. Ann Surg Oncol (2012) 19: 373. doi:10.1245/s10434-011-2051-2



Surgeons, along with the Centers for Disease Control and Prevention, emphasize the importance of managing symptoms and improving the quality of life of cancer survivors. A 2008 meta-analysis of mindfulness-based stress reduction (MBSR) concluded that this technique might improve patients’ adjustment to their disease. However, randomized controlled trials using standardized measures for evaluating MBSR are limited. The primary objective of this study was to evaluate, using valid and reliable measures, the effects of a unique, interactive, 8-week cancer recovery and wellness program on symptoms and quality of life of female cancer survivors.


Sixty-eight female cancer patients were randomized into either an intervention or waitlisted control group. Patients were evaluated using the Symptoms Checklist (SCL-90-R), the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-30), and the Symptoms of Stress Inventory (SOSI).


Of the participants, 70.6% were breast cancer survivors. Mean age was 57.5 years (treatment group) and 56.4 years (control group). Between-group demographic differences were not significant (P > 0.6). The treatment group improved significantly on the EORTC QLQ-30 (P = 0.005), on six of the eight SOSI subscales (P ≤ 0.049), and on both SCL-90-R subscales (P ≤ 0.023), while the control group did not improve on any of these measures (P > 0.2).


The MBSR-based cancer recovery and wellness intervention improved the symptoms and quality of life of this largely breast cancer survivor population across a variety of cancer symptoms and quality-of-life measures.

There are now more than 2.5 million breast cancer survivors in the USA. The relative survival of women with breast cancer at 5 years after initial diagnosis is 89%.1 The increased incidence and survival rates have resulted in over 11.7 million US cancer survivors, spurring major research into strategies for enhancing quality of life, reducing risk of recurrence, and extending survival.2,3

The diagnosis of cancer is a highly stressful experience that includes a cascade of distressing comorbid symptoms.4 Although increasingly efficacious for improving survival, current cancer treatments are debilitating in many ways, and they produce negative short-term and long-term psychological and physiologic effects including reduced quality of life, decreased cardiorespiratory capacity, pain, fatigue, and suppressed immune function.5 After cancer treatment ends, survivors continue to report high levels of psychologic stress, anxiety, depression, fear of recurrence, sleep dysfunction, impaired quality of life, and residual physical symptoms of pain and fatigue.6 Due to these significant comorbidities, from diagnosis through the balance of life, many cancer survivors seek information about food choices, physical activity, dietary supplement use, and integrative therapies for improving their response to treatment, quality of life, and survival.7,8 The Centers for Disease Control and Prevention (CDCP) emphasizes the importance of improving the quality of life of cancer survivors; emotional wellness and exercise are among the leading CDCP recommendations.9

The Silver Linings Cancer Wellness Program is a unique and interactive cancer recovery workshop that has been offered at our hospital since September 2005. The curriculum is geared toward female cancer survivors and includes the tools of mindfulness-based stress reduction (MBSR), which is a widely used stress management technique developed at the University of Massachusetts Medical Center by Kabat-Zinn et al.10 MBSR is a particular approach to stress management using mindfulness. Mindfulness is defined as “moment-to-moment awareness” while remaining nonjudgmental and accepting of different sensations, thoughts, and perceptions as they occur. In addition to MBSR, patients are taught mindful listening and writing, and the BrA method of breast awareness/self-examination. While controversy remains about the value of teaching breast self-examination, it is recommend for high-risk women by the National Comprehensive Cancer Network, and breast awareness is endorsed by a wide range of oncologic organizations and experts.1113

A 2008 meta-analysis of MBSR and cancer concluded that this method might improve patients’ adjustment to their disease.14 Even though evidence to date shows that MBSR methods may benefit survivors by decreasing symptoms and improving quality of life, randomized controlled trials using measures with known validity and reliability are limited. The primary objective of this study was to evaluate the effects of the unique, interactive, 8-week Silver Linings Cancer Wellness Program on female cancer survivors using standardized measures of symptoms related to cancer and its treatment, quality of life, stress, and psychological symptoms.


Sixty-eight female cancer patients 18 years of age and older were recruited to participate in the Silver Linings Cancer Wellness Program offered through the oncology department of a major, Mid-Western university teaching hospital over the proposed study period, September 2009 through August 2010. Patients throughout the hospital and metropolitan area were solicited, and providers were informed via personal contacts, pamphlets, websites, and other advertising about the opportunity for study participation. Potential participants were excluded if they had not completed their cancer treatment(s), or were judged during an entry interview to have significant psychiatric symptomatology that might impair their functioning in a group setting. Those who had planned or ongoing hormonal therapy, or who were on maintenance chemotherapy, were accepted if their treatment or disease was not expected to limit participation. Institutional review board approval for the ethical use of human subjects was obtained through the participating institutions. Subject characteristics are presented in Table 1.
Table 1

Subject characteristics


Treatment group

Control group

P value

Mean age in years (standard deviation)

57.5 (±10.5)

56.4 (±9.8)


Average years since diagnosis (standard deviation)

3.9 (±5.1)

3.7 (±3.5)


Subjects with breast cancer




n (%)

% of treatment group

% of control group





 White 56 (85.3)



 Black 8 (11.8)



 Asian/Pacific islander 2 (2.9)



 Missing data 2


Relationship status



 Single 8 (11.8)



 Married 44 (64.7)



 Living with partner 1 (1.5)



 Divorced 11(16.2)



 Widowed 4 (5.9)



Educational status


 High school/GED 9 (13.2)




 Some college 18 (26.5)




 College degree 13 (22.1)



 Postgraduate work 22 (32.4)



 Professional degree 4 (5.9)



 Missing data 2





 US $0–24,999 6 (8.8)



 US $25,000–49,999 8 (11.8)



 US $50,000–74,999 17 (26.5)



 US $75,000–99,999 21 (30.9)



 US $100,000 + 12 (20.6)



 Missing data 4


GED General equivalency diploma

After obtaining informed consent, patients were evaluated before and after the workshop series using standardized instruments that have published validity and reliability information and that have been cited extensively in the oncology and MBSR literature. The European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-30) is a 30-item standardized instrument which is considered the gold standard for evaluating quality of life for clinical trials in Europe and North America and which has been used in previous studies incorporating MBSR.15 Data revealed that the quality of life, function, and symptom scores were sufficiently interrelated and internally consistent to legitimately allow them to be averaged into a composite score. This single score was used to match and randomize patients to either the treatment or waitlisted control group (Fig. 1). Breast cancer patients completed the additional EORTC QLQ-30 BR23 questions that are geared to breast cancer-related quality of life.15 The Symptoms of Stress Inventory (SOSI) was used to evaluate physical, psychological, and behavioral responses to stress.16 This instrument has been used in other studies on cancer patients, yielding significant (P < 0.01) pre/post scores.17 The symptoms checklist (SCL 90-R) is a 90-item self-report symptom inventory which evaluates patients’ medical and psychological symptom patterns.18
Fig. 1

Participant flowchart

Treatment group patients were taught and practiced meditation, yoga, mindful communication skills, and mindful breast self-examination/awareness in weekly 2-h classes and at a single 4-h weekend retreat held during week 6 of the 8-week program. Patients also practiced daily meditation and/or yoga as a form of mild exercise, and they recorded daily logs that were used to verify adherence to the program. Sessions were conducted by the workshop leader, a physician who was formally trained in MBSR and teaching yoga. Class size maximum was 15. Each session began with a seated, 20-min meditation combining silence with verbal guidance in maintaining mindfulness. The ratio of silence to verbal guidance increased throughout the program. At each session, patients were provided the opportunity to “check in,” which involved a discussion of their experiences with the assigned, daily home mindfulness practices, and any other concerns, issues or problems they wanted to share with the group.

During the first session, patients were given an overview of the program’s goals. Each session included spacious listening, which comprises open, attentive listening without planning a response, passing judgment, or making assumptions. Week 2 included an introduction to the body scan, which is a guided meditation with attention focused on each body region. During week 3, patients experienced mindful eating and walking, and nonjudgmental acceptance and awareness of everyday activities.19 Yoga was introduced during week 4, and during week 5, patients learned spacious writing, which involves writing as thoughts occur for 10 min without concern for spelling or attention to logic.20 For week 6, patients were instructed to use a symbolic object to facilitate sharing of personal experiences with the group.21 Following week 6, a half-day retreat was held where participants practiced methods previously learned in the workshops. Week 7 included the BrA method of breast self-examination: a didactic presentation covering breast self-examination technique, followed by yoga which is used to lessen anxiety and promote thoroughness in the breast self-examination, and finally, guided practice of breast self-examination. The final week was reserved for discussion of body image and sexuality issues. Throughout the program, all subjects were requested to practice meditation techniques (sitting, walking, body scan, and yoga) for 45 min each day and to keep daily logs that were collected weekly. To avoid possible contamination with the intervention components, subjects who were randomized to the waitlisted control group were instructed not to use or practice meditation or yoga techniques.

Previously published studies have reported medium to large effect sizes expressed as Cohn’s d, where a d value of 0.20 is typically used to represent a “small” effect size, 0.50 a “medium” effect size, and 0.80 a “large” effect size. Generally, effect size represents the difference between a treatment group and the control group. Medium to large effect sizes are generally suggested for behavioral research.22 These published studies were able to detect statistically significant between-group differences using 49–77 patients.17,2325 Assuming a medium effect size of 0.5, a power level of 80%, and a two-sided alpha of 0.05, an estimated 64 subjects were required for the parametric variables. The EORTC QLQ-30 composite variable (described above) that was used to randomize participants was also found to satisfy the normality assumptions underlying the use of repeated-measures analysis of variance for statistically analyzing participants’ change scores. Analysis of responses to individual questions were performed to identify noteworthy changes among individual quality-of-life questionnaire items. These nonparametric variables were compared using the Mann–Whitney U test. Demographic characteristics were compared using the chi-square or Fisher’s exact test where appropriate. Throughout this study, two-sided P-values ≤0.05 were considered statistically significant. To include as many patients as possible in the treatment group, subjects were randomized 3:1 (treatment to control group); the statistical power calculation showed that this ratio preserved statistical power using the estimated 64 patients.


The major cancer type represented was breast cancer (70.6%) (Table 2). The number of diagnoses by cancer type among the treatment group (n = 48) and control group subjects (n = 20) did not differ significantly (P = 0.94), nor did they differ by age (years ± standard deviation, 57.5 ± 10.5 vs. 56.4 ± 9.8, respectively, P = 0.69), or years since diagnosis (3.9 ± 5.1 vs. 3.7 ± 3.5, respectively, P = 0.92). Treatment subjects attended an average of 6.9 ± 1.1 of the eight classes; thus weekly class attendance was 86.5%. Thirty-eight participants (79.2%) attended the half-day retreat. Two participants in each group reported having experience with the integrative practices used in this study. Of those with previous experience, none was a regular practitioner. Because the study interventions were applied in a prescribed and unique way, it was expected that prior experience of these participants would not affect the study results.
Table 2

Participants’ description by cancer type




Treatment group

 Unaffected BRCA+



 Breast cancer



 Ovarian cancer



 Endometrial cancer



 Hodgkin’s lymphoma



 Colon cancer






 Non-Hodgkin’s lymphoma



 Treatment group total



Control group

 Breast cancer



 Ovarian cancer



 Endometrial cancer



 Hodgkin’s lymphoma



 Colon cancer



 Non-Hodgkin’s lymphoma



 Control group total



Treatment plus control



Treatment and control groups did not differ significantly by cancer type (P = 0.94)

The intervention group improved significantly on the EORTC QLQ-30 (P = 0.005), on six of the eight SOSI subscales (P ≤ 0.049), and on both SCL-90-R subscales (P ≤ 0.023), while the control group did not improve on any of these measures (P > 0.2) (Table 3). These significant values represented mostly moderate effect sizes, with computed Cohen’s d values ranging from ±0.29 to ±0.56. Differences between the initial versus follow-up scores on these measures for the control group (n = 20) were not significant (P > 0.14). Among the subset of participants with breast cancer (n = 32), a moderate correlation (r = 0.42) was found between retreat participation and functional improvement on the EORTC QLQ-30. However, low correlations (r < 0.25) were found between the treatment group subjects’ hours of mindfulness practice, number of classes attended or retreat participation, and the change in any of their psychosocial assessments recorded among the study participants as a group.
Table 3

Mean difference and effect size of treatment group subjects’ pre- versus post-intervention assessments

Instrument and domain measured

Mean difference

P value

Effect size

EORTC QLQ-30 (n = 48)









 Symptoms (R)




 Overall score




SOSI (n = 48)









 Muscle tension




 Cardiac symptoms




 Sympathetic arousal








 Cognitive disorder




 Upper respiratory




SCL-90 (n = 48)









EORTC QLQ-30 BR-23 (n = 32)





 Symptoms (R)




 Overall score




Note Effect size (Cohen’s d for paired samples): ±0.20 = small; ±0.50 = moderate; ±0.80 = large. All effect size calculations are based on pre-program minus post-program assessments. All differences between the initial versus follow-up scores on these measures for the control group (n = 20) were not significant (P > 0.14)

aGlobal Severity Index (GSI) measures overall psychological distress

bPositive Symptom Distress Index measures intensity of psychological symptoms


Excellent improvements were evident among the treatment group patients compared with control group patients on almost every measurement assessed. Statistically significant improvements in the treatment group patients’ pre- versus post-treatment mean assessments were observed for ten of the 13 separate dimensions measured across the EORTC QLQ-30, SOSI, and SCL-90-R instruments, and one other score difference (cognitive disorder) approached significance (P = 0.052). Differences between the initial versus follow-up scores for the control group participants’ evaluations were all nonsignificant (P > 0.14). Because the treatment and control group patients’ average elapsed times since their cancer diagnosis were not significantly different, including this variable as another between-subjects effect in this analysis did not alter the conclusions.

It was not surprising that there was moderate correlation (r = 0.42) between retreat participation and functional improvement among the breast cancer patients. However, it is difficult to explain why this effect was not found among the treatment group overall. We may speculate that, because breast cancer was the most represented cancer type among the study participants (70.6%), they might have perceived the intervention to be more applicable to their particular illness. Perhaps the session on the BrA method had a more positive effect on breast cancer survivors versus survivors of other cancers. Popular support and hope for “the cure” for breast cancer might also have influenced these patients’ positive psychological disposition and level of adherence to interventions, including surgical and medical follow-up.

While previous randomized studies have shown efficacy for mindfulness training in improving symptoms and quality of life of cancer patients, in our study statistical interaction between the control and study groups’ score changes over the 8 weeks did not reach significance (P ≥ 0.145). Because many patients in the waitlist control group were planning to start the program soon after completing the follow-up instruments, this positive anticipation might have influenced their response.11,24 In our study, an average of 4 years and as long as 29 years had elapsed since cancer diagnosis. These longer-term survivors may respond differently to the intervention and/or anticipation of the intervention. In any event, the large number of long-term cancer survivors who seek wellness and treatment support programs suggests that many patients experience distress years after cancer treatments are completed. Long-term survivors may be well served by clinicians who inquire about their quality of life and symptoms and make appropriate referrals to survivorship resources.

For women who are cancer survivors, completing treatment can be almost as difficult as going through it. The support of regular clinical visits, the assistance of family and friends, and the reassurance of actively treating the cancer all fade away. Fear of cancer recurrence, anxiety, and depression are common. Patients need to regain a sense of trust in the world and in their body and to find their new “normal” as a posttreatment survivor. Many women are surprised to find they are not happy when treatment ends, but instead ask “what’s next?” Science saved their lives, but now it is up to them to ensure the quality and quantity of it moving forward. There is a gap between treatment ending and the discovery of “what’s next?” More research in this area is needed addressing the posttreatment period.

The present quantitative data are consistent with the qualitative data collected since the origin of the Silver Linings Cancer Wellness Program prior to the present study. Virtually every participant has recommended the program to other cancer survivors. Based on the study results, it was concluded that the MBSR-based Silver Linings Cancer Wellness Program improved the symptoms and quality of life of this largely breast cancer survivor population across a variety of measured symptoms and quality-of-life evaluations.


This study was funded in part by the 2010 Oakland University–William Beaumont Hospital Multidisciplinary Research Award. We gratefully acknowledge the study participants for their courage, determination, and openness.

Disclosure and Financial Support

The authors have no conflict of interest or commercial interests in the subject of study.

Copyright information

© Society of Surgical Oncology 2011