The trusted source for
healthcare information and
Abstract & Commentary
Source: Burstein HJ, et al. Use of alternative medicine by women with early-stage breast cancer. N Engl J Med 1999;340:1733-1739.
Objective: To survey subjects on their use of conventional therapies, alternative medical treatments, and health-related quality of life during the first 12 months after diagnosis of early-stage breast cancer.
Subjects: Cohort of 480 women in Massachusetts with newly diagnosed, Stage I or II breast cancer who participated in a two-year statewide study of how women choose treatment for cancer. Of the 1,532 women potentially eligible to participate in the study, hospitals and physicians granted consent to contact 1,158, of whom 795 (69%) agreed to participate. Then hospitals were asked to approve an add-on study about subsequent medical care and health-related quality of life, including use of alternative medicine (AM).
Administrative delay and lack of hospital consent reduced the number of eligible women from 795 to 657. Of these, 536 completed a questionnaire at three months, and 480 completed a questionnaire at 12 months. All study analyses are limited to the 480 (73.1% of eligible participants) who completed the 12-month survey. Five of the 480 respondents reported a recurrence of breast cancer.
The 480 participants who completed the 12-month questionnaire did not differ from the 177 eligible women who did not complete the surveys by age, marital status, menopausal status, breast cancer stage, or type of surgery (mastectomy or lumpectomy). Respondents had significantly higher levels of education and income, were more likely to be white, and had fewer coexisting illnesses.
Methods: Patients who consented to participate received mailed questionnaires timed to arrive at three and 12 months after the initial surgery. Validated survey instruments were used to assess physical and emotional health and quality of life. "Summary scores" for physical and mental health were done at baseline, but at three and 12 months, repeat "summary scores" plus five other survey instruments were used.
Alternative medical therapies were divided into two categories: "healing therapies," which required physical action on or exposure of the body (diet, massage, megavitamins, chiropractic, homeopathy, acupuncture, herbal remedies, energy healing, and folk remedies), and "psychological therapies," which involved primarily mental processes (imagery, hypnosis, relaxation, biofeedback, spiritual methods, and self-help groups). The use of alternative therapies was classified as "none," "new" (initiated within one year after breast cancer surgery), or "continuous" (used both before and after surgery).
Results: The use of alternative therapies was common after breast cancer surgery, with 135 of 480 women (28.1%) using some type of AM for the first time. New AM use accounted for over 70% of all use. The 135 new AM users used an average of 2.5 different alternative treatments. Most patients (71%) reported that their physicians knew of their AM use. Women with Stage II disease and those who received chemotherapy were significantly more likely to initiate AM use, whereas women who were taking tamoxifen were significantly less likely to do so.
The study reports that all groups of patients had similar baseline mental and physical health scores. Although physical health scores were similar at the three-month survey, the mental health scores of new AM users had fallen from the baseline, whereas the scores for non-users were higher than at baseline. On survey instruments initiated at three months, new AM users also reported more depression, less sexual satisfaction, greater fear of recurrence, and more numerous and severe somatic symptoms. At 12 months, all groups of patients had improved their mental and physical health scores, although new AM users still reported less sexual satisfaction and greater fear of cancer recurrence.
Comments by William Benda, MD and Charlea T. Massion, MD
Sometimes we find a study that is visionary in its implications but myopic in its conclusions. Burstein et al report that women with early-stage breast cancer who begin to use AM suffer more psychological distress, and therefore, new use of AM may identify the need for psychological, perhaps even psychiatric, intervention. However, another very plausible conclusion is that any woman faced with the loss of health and the specter of dying may experience extraordinary feelings of anxiety, depression, and vulnerability. The exploration of unconventional therapies and support from those practitioners may simply allow her to express emotions that are normal in the context of this intense health crisis.
In our time-stressed, managed care era, a patient is less likely to explore such territory with a physician than with a self-help group or a massage therapist. The "worsening" of mental health scores at three months may identify women who have the strength and willingness to confront their emotions and the implications of their disease early in the course of their treatment. It would be interesting to see the data beyond the 12-month completion of the study—perhaps those who chose alternative therapies eventually will suffer less psychological distress and experience better quality of life. This certainly was evident in Spiegel’s studies on improved breast cancer survival in women who participated in support groups.1 (Spiegel’s well-known and respected research is curiously absent from this study’s discussion and bibliography.)
The methodology must be called into question as well. There are several major problems with this study. First, although baseline summary scores for physical and mental health are reported, baseline scores are not provided for the five other survey instruments (the Medical Outcomes Study [MOS] sexual-satisfaction survey, the MOS 36-item short-form health survey, the Center for Epidemiologic Studies depression scale, the Lasry and Margolese scale for fear of breast cancer recurrence, and an expanded symptom checklist from the National Surgical Adjuvant Breast and Bowel Cancer Prevention Trial). Only three-month and 12-month scores are reported. Baseline differences in the scores of the two groups cannot be ascertained; the possibility exists that women who may have had lower baseline scores on these measures were more likely to have sought additional therapies, and therefore cannot be fairly compared to those with higher baseline scores.
Second, the differences in scores were appropriately analyzed with the Wilcoxon rank-sum test, but the distributions of scores were displayed as means. This is inappropriate and prevents the reader from judging whether there were clinically important differences among women in the various groups. Just a few outliers can throw off the mean for the entire group.
Also, those who chose to participate in the survey were younger, more often white, had higher income and educational levels, and had fewer coexisting illnesses than the eligible non-participants. Based on previous research, white women with a higher socioeconomic status are more likely to seek alternative therapies anyway. This calls into question the authors’ conclusion that women who are new AM users are more prone to psychological dysfunction. The truth may be that they are simply more likely to discuss their fears and feelings based on their prior experiences and behaviors. Thus, there may be an inadvertent, yet significant selection bias.
The above criticisms do not necessarily negate the validity of the overall study. However, to assume that seeking help which is generally unavailable in conventional medicine denotes psychological pathology rather than potential resilience reveals a disturbing biomedical bias, and offers troubling implications if the authors’ "findings" change clinical practice without further research. In an accompanying editorial, Sloan Kettering’s Dr. Jimmie Holland urges physicians to ask about alternative therapies to screen for women who need psychiatric consultation. In response, one breast cancer activist stated, "If physicians were to follow Dr. Holland’s advice to consider new AM users for psychiatric referral, particularly in our age of managed care, women would be unlikely to encounter a sympathetic listener, but would probably be given a tranquilizer or antidepressant prescription, possibly soothing symptoms, but unlikely to deal with underlying fears or upsets."2
It’s true that women with breast cancer are at higher risk for anxiety and depression and that their physicians need to be alert to these symptoms. It’s also true that many physicians need to improve their screening for these conditions. One study of oncologists found that they markedly underestimated depressive symptoms in their most depressed cancer patients.3 But why screen for depression by asking about AM use? DSM-IV describes a simple and inexpensive checklist for depression screening, or you may use the Beck, Zung, or other depression rating scales.
Perhaps further observation and evaluation of the new AM users could provide clues to help breast cancer patients become breast cancer survivors. Conceivably, AM use is a marker after all, but one of emotional health rather than psychopathology. In any case, the methodological problems with this study should be considered when judging its conclusions.
Dr. Benda is a Senior Fellow in the Program in Integrative Medicine at the University of Arizona in Tucson, AZ; Dr. Massion is on the editorial advisory board.
1. Spiegel D, et al. Effect of psychosocial treatment on survival of patients with metastatic breast cancer. Lancet 1989;2:888-891.
2. Jane Zones. Breast Cancer Action. San Francisco, CA: Breast Cancer Action. In press.
3. Passik SD, et al. Oncologists’ recognition of depression in their patients with cancer. J Clin Onc 1998;16:1594-1600.