Background
Early detection and improved multimodality therapy have improved breast cancer survival [
1]. When first diagnosed, patients with breast cancer must make decisions about the type of surgery and adjuvant therapy, each with its own risks and benefits. The decision-making process is complex, and is influenced by multiple factors including patient age, co-morbidities, ethnicity, education, and availability of visual and written hand-outs. Patient age is so important that it has been proposed as a determinant of educational and counseling strategies [
2,
3].
Age at diagnosis may also influence patient behavior and outcome. Older women prefer to take a more passive role in decision-making, relying heavily on their doctors’ recommendations [
3], whereas younger women are more actively involved in treatment decisions, seeking information from multiple sources such as the Internet, friends, family members, and support groups. It has been found that patients provided with a greater amount of medical information were associated with less depression and improved quality of life [
4]. In the same study, patients who were satisfied with their participation in the decision-making process were also found to be less depressed. Younger women may desire treatments that they perceive increase life expectancy, while older women may prefer treatments that maximize quality of life [
5,
6].
The current study was a retrospective analysis of treatment selection and perceived quality of life in a cohort of breast cancer survivors. We hypothesized that age at diagnosis of breast cancer significantly influenced patient selection of type and duration of therapy, independent of extent of disease, and that it was directly related to perceived quality of life after treatment. A cross-sectional and correlational study design was used.
Methods
The Comprehensive Cancer Center of Rhode Island Hospital is a tertiary cancer center that treats approximately one-third of all patients diagnosed with breast cancer in Rhode Island. All breast cancer cases are presented prospectively at the multidisciplinary breast tumor board and recorded in the Rhode Island Hospital Breast Cancer database. Patients with breast cancer diagnosed from January 2004 to June 2007 were identified from this database. Female patients with invasive breast cancer or ductal carcinoma in situ without known recurrent disease were included. Male patients were excluded.
A survey evaluating breast cancer treatment and quality of life was mailed to patients with breast cancer on 15 July 2008 (Additional file
1). Patients were asked to complete the survey and return it anonymously in a pre-stamped addressed envelope. A raffle was used to increase survey participation, but the raffle form was returned in a separate pre-stamped addressed envelope to protect the anonymity of participants. Both English and Spanish versions of the survey were available. The English version of the survey was mailed initially with information on how to obtain the Spanish version included at the top of the survey. This study was approved by the Rhode Island Hospital Institutional Review Board.
The European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30 [
7] and BR23 [
8] questionnaires were utilized in the survey. The QLQ-C30 module includes global health, functional, and symptom subscales which are robust survey tools for psychometric evaluation and reproducible across population [
7]. The BR23 module evaluates functional and treatment-related symptoms specific to breast cancer, including body image questions. Additionally, the survey included 56 independently developed questions (supplementary section) related to cancer characteristics (4), demographics (10), extent of disease (2), use and type of surgery (6), radiotherapy (6), chemotherapy (7), anti-hormones (5), support systems (10), and perception of therapies (6). Decisions about treatment were collected by asking patients to identify the surgery their clinician recommended, the reason for the recommendation, if known, and the surgery chosen. The survey also asked participants to identify whether radiotherapy, anti-hormone medication, or chemotherapy was recommended, and their final treatment selected. Patients were provided a list of personal, social, and professional support structures, and were asked to select the three which most influenced their treatment decision. Eight questions assessed how patients felt about their treatment choices and perceived quality of life. The Decision Regret Scale, a validated five-item scale for measuring regret after treatment choices, was used to assess patient satisfaction with treatment selected [
9]. Perceived quality of life was measured by the EORTC QLQ-C30 [
7] and the QLQ-BR23 [
8], and included questions about fear of recurrence and death.
We used the average age at onset of menopause (50 years) and the qualifying age for Medicare benefits (65 years) to divide patients into three age groups: young (<50 years), middle-aged (50-65 years), and old (>65 years). Extent of disease (local represented disease confined to the breast and regional indicated disease that had spread to regional lymph nodes) and treatment differences were analyzed for each age group. Correlations between quality of life and social support, decision regret, education, and income were determined. Raw scores were calculated according to the Likert scale of individual surveyed items. Linear transformation was used to convert raw scores of global health status and functional scales to a scale of 0 to 100, with 100 being highest level of function and quality of life. Symptom scales (examples: lymphedema, hair loss) were scored from 0-100, with 100 being most symptomatic. Future perspective of patients was scored from 0-100, with 100 being most optimistic about their health and prognosis. The Decision Regret Scale ranged from 1-5, with 5 being strongly agree.
Statistical analysis
A chi-square test was used to compare categorical results and variables across age groups. Fisher’s exact test was used to calculate P-values when 20% of the cells had values below 5. For continuous variables, analysis of variance (ANOVA) or Kruskal-Wallis tests for three or more groups (when the data were skewed) were conducted. Logistic regression models were used for continuous predictor variables with dichotomous outcome variables. The statistical significance was set at P < 0.05. Data were analyzed using SAS version 9.2 software.
Discussion
Breast cancer therapy consists of multimodality treatment ranging from breast conservation to mastectomy, whole or partial radiotherapy of varying duration, selective use of anti-hormone medication, and chemotherapy regimens that differ in duration and morbidity. Our data supported the original hypothesis that age is related to treatment choice; young women were more likely to choose aggressive therapies, even when diagnosed with localized disease. This result is consistent with the observation that young women may select more aggressive treatments in order to maximize their survival [
11,
12]. By contrast, Bleicher and colleagues [
3] reported no difference in mastectomy choice by age, and a study from the University Hospital of Wales reported that older patients were more likely to choose mastectomy as a treatment option [
13].
Increased awareness of family history may increase the likelihood of young patients pursuing mastectomy. Although these variables were not included in our survey, it has been reported that women with a family history of breast cancer were more likely to choose mastectomy as their breast cancer treatment option [
14]. Women with a family history of breast cancer are also more likely to undergo prophylactic mastectomy [
15]. This may partially explain why young women in our study were more likely to undergo more invasive surgery, including prophylactic mastectomy, because the risk of recurrence and desire for longevity are significant patient concerns.
The need for maximal survival may influence surgical treatment choice [
16]. Patients with breast cancer may opt for mastectomy, as they perceive that this procedure increases cancer survival. Although local recurrence decreases with mastectomy, breast-conserving surgery is no less efficacious than mastectomy with regard to long-term survival [
16]. In general, a higher education level was associated with opting for breast conservation, which may be due to a better understanding of risk recurrence (Surveillance, Epidemiology, and End Results (SEER) Program, Fast Stats-Breast Cancer). However, young patients in our study were better educated but less likely to choose breast-conserving surgery. The increased use of mastectomy by young women may be a reflection of feeling comfortable with invasive surgery due to having a greater social support network of family and friends. In one prior study, elderly breast cancer survivors reported having less social support network mechanisms than young women [
17]. In the population studied in this report, an increase in support group utilization across all age groups may present a future research direction.
In addition to choosing less invasive surgery, older women were also less likely to receive chemotherapy. These women were less likely to use anti-hormone medication despite their physicians’ recommendations. Furthermore, older women tend to have poorer general health than young women, and are more likely to have medical co-morbidities [
17]. This may, in part, explain why more older women chose not to receive chemotherapy or anti-hormones as part of their breast cancer treatment.
In our study, physician recommendation was of primary significance in influencing the patient’s selection of treatment. Physicians may be less likely to recommend medical interventions to older patients because of potential complications and increased morbidity [
18]. The authority for decisions regarding treatment may not be equally shared between physicians and patients. Physicians may also underestimate the influence of the words they choose to describe extent of disease and potential treatments options to patients [
19].
Overall, patients in this study reported being satisfied with information regarding treatment options. In addition, 83% of survey participants stated that their surgeon’s recommendation was a major influence on treatment selection options. A previous study which reviewed factors influencing surgery choice in women also showed similar results [
20]. Another study showed that 95% of study participants were satisfied with time spent discussing treatment options with their surgeons, underscoring a surgeon’s influence on the type of breast cancer surgery selected [
21]. The congruency between the patients’ and physicians’ opinions has been associated with treatment satisfaction, which in turn correlates with higher levels of quality of life [
22]. Similarly, another group found that the more the patient was involved in decision-making, the better the quality of life that she seemed to achieve by association [
23,
24]. These results reiterate the importance of physician’s participation in empowering and educating patients with breast cancer regarding their treatment options.
Most breast cancer survivors reported preserved quality of life and function in this study. Breast cancer survivors’ quality of life improves with time and is comparable to the general population five years after treatment [
25]. Over longer time periods, most breast cancer survivors have functional scales comparable to those of the general population [
26,
27]. Interestingly, patients in this study had a good quality of life after only an average of 31 months after diagnosis, suggesting that a return to function occurs much earlier than previously reported. There was no difference in side effects between the different age groups but, after controlling for extent of disease, age was significantly related to arm symptoms. In other studies, quality of life was lower in breast cancer survivors who experienced arm symptoms, with or without lymphedema [
28]. This might be expected in younger women who undergo more aggressive surgery and/or are more physically active and therefore more likely to notice functional limitations.
Poorer body image in young breast cancer survivors may reflect increased selection of mastectomy as a treatment option [
29]. Concerns for appearance have led younger patients to opt for lumpectomy more frequently compared to their older counterparts [
30]. It has also been reported that young patients significantly experienced more worries regarding their finances and future health, and had poorer emotional and social functioning compared to the older generation [
31]. Although socioeconomic status had no correlation with treatment decisions or self-perceived health-related quality of life, factors such as annual household income, disability status, and insurance coverage options may be barriers to health care access and consequently outcome in some communities [
32].
There are several limitations to this study. The conclusions of the study are based solely on patient responses and may reflect a recall bias. Our survey tools were based on the EORTC QLQ-C30 and QLQ-BR23 questionnaires, which do not evaluate social and spiritual quality of life [
33]. Although the survey was generated in an academic urban center, the vast majority of respondents were Caucasian with private health insurance. Thus, future studies which incorporate minority populations will be desirable, since cultural belief may represent a major factor in medical decision-making.
The survey response sample data set in our study was self-selected. Although the survey was mailed to all patients who were treated between 2004 and 2008, those who responded to the survey might have had better quality of life and health outcomes than those who chose not to respond. Breast cancer survivors may report fewer symptoms and better quality of life if they have not had a recurrence or a therapy-related complication [
34]. This may have introduced a non-random sampling bias into the study results. Characteristics of survey participants and non-participants could not be compared because of the anonymous nature of the study.
Perceived quality of life may differ based on ethnicity [
35] or on whether patients had private health care insurance. The initial survey was written in an English language format, which may have decreased the survey response rate of patients who speak English as a second language or not at all. A Spanish language survey format was also available to patients but only upon request.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
TTS, DW, MP, and MC designed the study, collected the data, and drafted the manuscript. TTS, DW, KC, RJ, and DM performed the literature review. TTS, KC, RJ, MP, and JS performed the statistical analyses. KC, RJ, DM, and JS also participated in the drafting of the manuscript. All authors read and approved the final manuscript.