Relationships Between Three Beliefs as Barriers to Symptom Management and Quality of Life in Older Breast Cancer Survivors

Abstract

Purpose/Objectives

To describe relationships among perceived barriers to symptom management and quality of life and to test the mediating role of perceived communication difficulties on the relationships between other perceived barriers to symptom management and quality of life in older breast cancer survivors.

Design

Cross-sectional descriptive-correlational design using baseline data from a randomized controlled trial that tested the efficacy and durability of an Individualized Representational Intervention (IRIS) in reducing symptom distress and improving quality of life in older breast cancer survivors.

Setting

Participants were recruited from the community, an oncology clinic, and a state tumor registry.

Sample

190 older breast cancer survivors (M age = 70.4) who were an average of 3.3 years after breast cancer diagnosis and who reported an average of 17.8 symptoms

Methods

Path analysis was conducted using Mplus.

Main Research Variables

Negative beliefs about symptom management (SMBQ), perceived negative attitudes from health care providers (CommA), and perceived communication difficulties (CommD), mental quality of life (MCS), physical quality of life (PCS), purpose in life (PIL), and positive relation with others (PR)

Findings

Significant direct effects of SMBQ and CommA on CommD were found after controlling for age, number of health problems, and number of symptoms. CommD was a significant mediator of the effects of CommA on quality of life, after controlling for the covariates. SMBQ had significant total effects on MCS, PIL, and PR after adjusting for the covariates, but was not mediated by CommD.

Conclusions

The mediating role of perceived communication difficulties with health care providers about symptoms suggests that patient-provider communication is an important factor in quality of life of older breast cancer survivors.

Implications for Nursing

Developing and testing nursing interventions focusing on enhancing both positive beliefs about symptom management and effective communication in old age is suggested.

Introduction

Older breast cancer survivors are a rapidly growing segment of cancer survivors. Women over the age of 75 have the highest incidence rates for breast cancer, with an overall survival rate of 85% (Altekruse et al., 2009). Thus, given the aging population, breast cancer will be one of the most prevalent chronic illnesses in older women (Ceber, Turk, Soyer, Ciceklioglu, & Cimat, 2006).

Effective management of symptoms is a critical component of the quality of life of older cancer survivors (Deimling, Bowman, & Wagner, 2007; Heidrich et al., 2009; Heidrich, Egan, Hengudomsub, & Randolph, 2006). Experiencing and managing symptoms as older cancer survivors is challenging because symptoms can be due to late effects of cancer and cancer treatment (Deimling et al., 2007) as well as the physiological declines of normal aging and age-related health problems (Heidrich et al., 2009). However, little attention has been paid to identifying or explaining the specific symptom management issues faced by older breast cancer survivors.

Beliefs about health problems, such as symptoms, play a role as “a filter and interpretive schema” that drives behaviors in response to the problems (Hagger & Orbell, 2003). Beliefs are affected by socio-cultural context (Leventhal, Brissette, & Leventhal, 2003), and thus beliefs about aging, particularly age-related stereotypes, may also serve as an interpretive schema through which symptom management of older people is filtered. Ageist stereotypes are defined as negative beliefs, discriminatory attitudes, and norms about older individuals or groups because of their age (Nelson, 2002). Empirical evidence indicates that older people share ageist stereotypes (Nussbaum, Baringer, & Kundrat, 2003; Morgan, Pendleton, Clague, & Horan, 1997). For example, older cancer survivors attributed many of their symptoms to aging rather than to cancer (Heidrich et al., 2006; Royer, Phelan, & Heidrich, 2009). When beliefs about aging are negative, they have detrimental effects on the older person’s health outcomes, such as physical and psychological health (Levy, Ashman, and Slade, 2009; Sánchez Palacios, Trianes Torres, & Blanca Mena, 2009; Yeom & Heidrich, 2009) and longevity (Levy, Slade, Kunkel, & Kasl, 2002).

Age-related stereotypes may be shared by health care providers (HCPs), in part due to their experience with fragile older patients (Greene & Adelman, 2003). Empirical studies have reported that when HCPs interact with older persons they tended to talk less, to provide oversimplified information (Nussbaum et al., 2003; Siminoff, Graham, & Gordon, 2006), and to belittle their concerns (Dawson, Sellers, Spross, Jablonski, Hoyer, & Solomon, 2005). These communication attitudes of HCPs have a negative effect on older persons’ health and quality of life (Mandelblatt et al., 2003; Yeom & Heidrich, 2009).

On the other hand, when communication between HCPs and older patients is positive, research has found better adherence to medical treatment (Mandelblatt et al., 2003), higher satisfaction with care (Stewart, Meredith, Brown, & Galajda, 2000) and higher overall quality of life (Liang et al., 2002; Stewart, et al., 2000). Communication with HCPs about symptoms is an important aspect of effective symptom management in that it provides the opportunity for either reinforcing or weakening barriers to symptom management.

The aims of this study were to investigate 1) whether negative beliefs about symptom management and perceived negative attitudes from HCPs are related to more perceived difficulty communicating about symptoms in older breast cancer survivors and 2) whether the perception of communication difficulties mediates the influence of either negative beliefs about symptoms or perceived negative attitudes from HCPs on quality of life in older cancer survivors.

Conceptual framework

A conceptual model of this study was developed based on the Leventhal’s Common Sense Model (CSM) and empirical evidence about aging-related beliefs and health outcomes in older people. The CSM is based on the idea that individuals have common-sense beliefs about a health threat that are shaped by the broader socio-cultural context, by interactions with others (e.g., health care providers), and by experience. These beliefs guide an individual’s coping behaviors in response to the health threat, which, in turn, influence health outcomes, such as quality of life (Leventhal, Meyer & Nerenz, 1980). Beliefs older people have about experiencing and managing symptoms (in this case related to aging and breast cancer) are formed in part by ageist stereotypes in the broader culture (see Figure 1). A study by Yeom and Heidrich (2009) found that aging-related beliefs of older cancer survivors were related to poorer quality of life outcomes, suggesting that these beliefs might work as perceptual barriers to symptom management.

Figure 1.

A conceptual model of the hypothesized relationships among three beliefs as barriers to symptom management and quality of life

The specific beliefs investigated in this study were: (1) negative beliefs about symptom management, defined as negative or stereotyped beliefs about experiencing or managing symptoms in old age; (2) perceived negative attitudes from HCPs, defined as negative and unsupportive attitudes of HCPs that older people have perceived in communicating with their HCPs; and (3) perceived communication difficulties, defined as the difficulties that older people perceive in knowing when and how to report symptoms to their HCPs. According to this model, negative beliefs about symptom management and perceived negative attitudes of HCPs lead to difficulties in communicating with HCPs about health problems. Not being able to communicate about symptoms results in less effective symptom management leading to lower quality of life. Thus,perceived communication difficulties mediate the influence of the other two beliefs on quality of life in older cancer survivors.

Background

Ageist beliefs about health

A number of studies have reported negative and sometimes erroneous beliefs about health and aging that are held by older people. Older people, including cancer survivors, tend to attribute their health problems to normal aging and to report that symptoms in old age are inevitable and incurable and therefore without a need for a medical evaluation for possible causes (Heidrich et al., 2006; Morgan et al., 1997). Royer and colleagues (2009) found that older breast cancer survivors described their symptoms as incurable and not controllable. Dawson et al. (2005) compared older and younger cancer patients and found that older patients were more likely than younger patients to believe that medication could not control their pain. A comparative study of symptom beliefs between older women who had breast cancer and those without a cancer history found that older women, regardless of breast cancer history, attributed their symptoms to aging rather than cancer or other diseases (Heidrich et al., 2006).

Negative beliefs about aging may have detrimental effects on management of health problems in older populations. Older people who view functional declines as a part of the normal aging are less likely to seek care (Sarkisian, Hays, and Mangione, 2002), to engage in preventive behaviors such as exercise and healthy diet, or to adhere to medical treatments (Levy & Myers, 2004).

Negative beliefs about aging and attitudes of HCPs

Health care providers are not free from ageist stereotypes (Nussbaum et al., 2003) and may even be more vulnerable to ageist stereotypes due to their increased exposure to fragile older patients (Greene & Adelman, 2003). Some examples of beliefs related to ageist stereotypes of HCPs are: older adults are too old to try a new treatment; older adults are unwilling to try new treatments; and treatment may be ineffective in older people (Kagan, 2008; Morgan et al., 2003). Health care providers also tended to regard caring for older patients who have typical aging-related symptoms as less rewarding than caring those who have disease-based symptoms (Baker, 1984).

Previous research has indicated negative attitudes of HCPs in communicating with older patients. Health care providers talked less with older patients and provided oversimplified information to them compared to younger patients (e.g., Nussbaum et al., 2003; Siminoff et al., 2006). In addition, older patients reported that they had been told that they “should be thankful to be alive” or “just learn to live with it” (Beisecker, 1988; Cegala, Post, & McClure, 2001). In addition, previous studies have supported that ageist stereotypes of HCPs could work as a barrier to effective communication about symptoms. Armstrong-Esther et al. (1989) found that nurses with more negative perceptions about aging were likely to undervalue the importance of communication in caring for older patients, whereas those with more positive views about aging placed more importance on having conversations with older patients than with only providing basic care such as bathing and toileting (Armstrong-Esther, Sandilands, & Miller, 1989).

Communication Difficulties

Communication with HCPs is an important route for older patients to interpret symptoms, to shape beliefs about symptom management, and consequently to determine further reactions to symptoms (Nussbaum et al., 2003). However, ageist self-stereotypes, erroneous beliefs about being a “good patient”, or unsupportive attitudes of HCPs may act as barriers to older cancer patients’ reporting symptoms. Older people explained the reasons for not reporting their health problems to HCPs as “no big deal,” “nobody cares,” “nothing can be done about it,” and “don’t want to bother people” (Stoller and Forster, 1994; Brody & Kleban,1981). Similar attitudes have been found in cancer patients, such as “a good patient does not talk about their symptoms” and “reporting non-cancer related symptoms may distract HCP” (Gunnarsdottir, Donovan, Serlin, Voge, & Ward, 2002).

There are empirical findings that support the importance of communication for better health outcomes in older patients. A study by Dawson et al. (2005) showed that, regardless of the sources of pain (cancer or non-cancer related), older patients’ willingness to take pain medications was dependent on whether or not they had communicated with HCPs about fears of addiction from pain medications. Mandelblatt et al. (2003) found that whether or not older patients communicated with HCPs about their pain was related to their satisfaction with care, and HCPs’ attitudes were more predictive of patients’ quality of life than the treatment they received.

Yeom and Heidrich (2009) examined the relationship between three beliefs as barriers to symptom management, and quality of life in 61 older breast cancer survivors (M age = 71). There were significant positive correlations (r’s ranging from .40 ~ .47) among negative beliefs about symptom management, perceived negative attitudes of health care providers, and perceived communication difficulties about symptoms. In regression analyses, negative beliefs about symptom management was a significant predictor of depression (β = .25, p < .05), and both perceived negative attitudes from HCPs and perceived communication difficulties were significant predictors of mental quality of life (β = −.31, p < .05; β = −.26, p < .05, respectively), after adjusting for the number of health problems and the degree of symptom bother. However, further investigation of multivariate relationships among the three beliefs and quality of life was limited due to the small number of subjects.

These studies suggest that older persons’ ageist self-stereotypes influence their beliefs about their symptoms and whether or not they attempt to manage or seek care for their symptoms. Furthermore, ageist stereotypes held by HCPs may influence how they communicate with older patients about their health problems and the treatment that is provided. Both may lead to difficulties of older patients in communicating with HCPs about their health problems. Difficulties communicating with HCPs may result in less effective symptom management because erroneous and stereotyped beliefs about aging can be either weakened or strengthened during health care encounters. The result is poor symptom management and lower quality of life for older persons.

This study examined two hypotheses: 1) negative beliefs about symptom management and perceived negative attitudes from HCPs will predict perceived communication difficulties about symptoms in older breast cancer survivors; and 2) the effects of negative beliefs about symptom management and perceived negative attitudes from HCPs on quality of life will be mediated by perceived communication difficulties.

Method

Design

This study was a descriptive correlational study. A secondary data analysis was conducted using baseline data from a randomized controlled trial testing the efficacy and durability of an Individualized Representational Intervention (IRIS) in reducing symptom distress and improving quality of life in older breast cancer survivors.

Sample and Procedure

To be eligible for the parent study, women had to be aged 65 or older, at least 1 year post diagnosis of local or regional breast cancer, at least 3 months post-treatment (surgery, chemotherapy, radiation therapy, but could be taking hormonal therapies), without metastatic cancer or recurrence of breast cancer or other cancer diagnosis (except non-melanoma skin cancer), and without cognitive impairment or physical frailty that could preclude participation.

The Health Science Institutional Review Board approved the parent study. Participants were recruited from multiple sites including the state tumor registry, the community, and oncology clinics in a Midwest area. Baseline measures were collected by mail and at a baseline telephone interview. The sample for the parent study was 190. A power analysis for this study with alpha as .05 and a small effect size (.2) was .99 using the LISPOWER (Joreskog & Quiroga, 1988) program.

Measures

Demographic, Health, and Breast Cancer Information

Respondents completed a demographic questionnaire asking age, education, income, ethnicity, marital status, and living arrangements.

Health status was assessed with three measures: the numbers of health problems, symptoms, and medications. The number of health problems was assessed with the 20-item self-report checklist adapted from the Wisconsin Longitudinal Study of Families-Illness and Medical Conditions. Respondents reported whether or not they had been told by HCPs that they had each of 20 health problems and the total number of the health problems was computed.

The number of symptoms was assessed with the 37-item Symptom Bother-Revised (SB-R) (Heidrich et al., 2006). The instrument includes 37 symptoms common to aging, age-related health problems, and late effects of cancer and its treatment. Whether or not a respondent had each symptom (i.e., 0 = “don’t have” and 1 = “have”) was assessed, and the total number of symptoms was calculated.

Respondents were asked to list names of their prescription and over-the-counter medications, and the total number of medications was calculated.

Information about breast cancer history included months since diagnosis, types of cancer treatment (i.e., mastectomy, lumpectomy, radiation, chemotherapy, lymph node removal/dissection, sentinel node biopsy), and past and current hormonal therapies, and months since the last cancer treatments.

Beliefs as Barriers to Symptom Management

Three beliefs that may act as barriers to symptom management were measured with three instruments: Symptom Management Beliefs Questionnaire (SMBQ), Communication Attitudes (CommA) and Communication Difficulty (CommD). In previous studies, all three measures were found to be reliable (Kuder-Richardson Formula 20 [KR-20] > .70), and construct validity was supported by significant intercorrelations among the three scales as well as significant correlations with measures of perceived health and quality of life (Heidrich et al., 2009; Yeom & Heidrich, 2009).

The SMBQ is a 13-item scale to assess the extent to which an older respondent holds negative beliefs about experiencing or managing symptoms (Heidrich et al., 2009). The items were developed by adapting some items from Ward’s Barriers Questionnaire II (Ward, Carlson-Dakes, Hughes, Kwekkeboom, & Donovan, 1998), which has been used to measure barriers to analgesic use for cancer pain, and by including items reflecting negative stereotypes about aging. Respondents rated the extent to which they agree with each item on a five-point Likert-type scale, from 0 (do not agree at all) to 5 (agree very much). A mean score was computed and higher scores indicate more negative beliefs about managing symptoms. Internal consistency in this sample was Cronbach’s alpha = .80.

The CommA is a 10-item dichotomous scale (yes = 1, no = 0) measuring the extent to which an older respondent has perceived ageist remarks or attitudes from their HCPs. The items were developed from the empirical evidence on ageist stereotypes and from pilot studies about adults and older cancer survivors (Heidrich et al., 2009). The original CommA (Yeom & Heidrich, 2009) had seven items that were focused on breast cancer issues such as “compared to other women with breast cancer” and “A lot women with breast cancer.” Based on results of another study (Heidrich, Brown, et al., 2009), three new items were added that were not breast cancer specific but were frequently mentioned regarding communication with HCPs. They were: “Feeling a symptom you reported was not dealt with to your satisfaction,” “Being told, not much can be done about one or more of your symptoms because they’re part of getting older,” and “Being told that you’ll have to learn to live with it because your symptom(s) are part of getting older.” The CommA with the three new items was reviewed for content validity by three adult and geriatric nurse practitioners. A sum score of 10 items was computed, and higher scores indicate that respondents have perceived more negative attitudes from HCPs. Internal consistency (KR-20) for the 10-item dichotomous scale in this sample was KR-20 = .78.

The CommD is a 6-item dichotomous scale (yes = 1, no = 0) that assesses the perception that communicating with HCPs about symptoms is difficult. This scale was developed from the literature on ageist stereotypes and from pilot studies with adults and older cancer survivors (Heidrich et al., 2009). A sum score of 6 items was computed and higher scores indicate that respondents perceive more difficulties communicating with HCPs about their symptoms. Internal consistency (KR-20) for the 6-item dichotomous scale in this sample was KR-20 = .80.

Reliability and validity of all three instruments has been reported (Heidrich et al., 2009; Yeom & Heidrich, 2009). Construct validity of the SMBQ, CommA, and CommD was examined by testing the associations among these three measures. The three scales should be correlated because all three tap perceptions and beliefs about symptom management. However the relationships should be small to moderate because each scale taps a different type of belief or perception. In prior studies, this pattern of correlations was found: between the SMBQ and CommA (r = .25 ~.40; p < .05), between the SMBQ and CommD (r = .32 ~.44; p < .05), and between CommA and CommD (r = .32 ~.47; p < .05).

Quality of life

In this study, quality of life was conceptualized broadly in order to capture the important dimensions of quality of life for older adults, which include physical quality of life, mental quality of life, purpose in life, and positive relations with others.

Physical quality of life and mental quality of life were measured with the Physical Component Summary (PCS) and Mental Component Summary (MCS) scores of the SF-36 scale (Ware & Sherbourne, 1992). The PCS and MCS scores were computed by using specific scoring instructions resulting in standardized scores ranging from 0 –100, and higher scores indicate higher physical functioning, mental functioning, and quality of life. The SF-36 has been used in a number of clinical trials with breast cancer patients as a quality of life measure. The validity and reliability of both the PCS and MCS have been demonstrated in numerous studies in both general and disease-specific (including cancer) populations (Schlenk et al., 1998). The reliability in this study was α = .93 (PCS) and α = .95 (MCS).

Purpose in life and Positive Relations with Others are two of six core dimensions of psychological well-being proposed by Ryff (1989). The dimensions were derived from adult development theory and empirical evidence about their significance for physical and mental health outcomes in the elderly population (Ryff & Keyes, 1995).

Purpose in life refers to having a sense that one’s life is meaningful and purposeful even under threats (Ryff, 1989). Purpose is related conceptually to spirituality and to “finding meaning”, both of which are related to adaptation to breast cancer (Tomich & Helgeson, 2002). Purpose in life was measured with the 14-item PIL scale, and respondents were asked to respond how strongly they agreed or disagreed with each statement as a 6-point scale ranging from 1 (strongly disagree) to 6 (strongly agree). Negative items were reverse coded, and a mean score of the 14 items was computed. Higher scores indicate higher levels of a sense of purpose in life.

Positive relations with others is characterized as having warm and trusting interpersonal relations and being concerned about others’ welfare (Ryff, 1989). Positive relations with others was measured with the 14-item PR scale, and respondents were asked to respond how strongly they agreed or disagreed with each statement as a 6-point scale ranging from 1 (strongly disagree) to 6 (strongly agree). Negative items were reverse coded, and a mean score of the 14 items was computed. Higher scores indicate that people have satisfying and trusting relationships with others.

The validity of PIL and PR have been demonstrated by significant relationships with other indices of psychological well-being in a number of cross-sectional, longitudinal, and cross-cultural studies in older populations (e.g., Schnoll, Knowles, & Harlow, 2002; Tomich & Helgeson, 2002). Both scales are reliable (alphas range from 0.86 to 0.93; test-retest reliability over 6 weeks ranges from 0.81 to 0.88) (Ryff & Keyes, 1995). The reliability of the PIL and PR in this study was α = .88 and α = .88, respectively.

Data Analysis

Preliminary analyses and descriptive statistics were conducted using SPSS 16.0 to describe the sample, measures, and correlations among the main study variables.

Hypotheses were examined with path analysis using Mplus 5.1 (Muthen & Muthen, Los Angeles, CA). The CommD values that were measured with dichotomous variables were transformed to an appropriate form for multivariate testing using square root transformation. The measurement errors for the composite variables were addressed by utilizing the Composite Indicator Structure Equational Modeling (CISE), which is a way to make the estimate more accurate by adjusting for measurement error of a composite scale (Bandalos, 2002).

Results

Description of the Sample

Table 1 shows the demographic characteristics, health status, and breast cancer history of participants. The mean age of the participants was 70.4. The average year of education was 14.2, and over half of the participants had annual income over $40,000. The majority of the participants self-identified as Caucasian (n = 178, 93.7%), and a small minority of participants identified as Black or African American (n = 8, 4.2%). This reflects the population in the Midwestern area where recruitment took place. The majority of the participants (n = 111, 58.4%) were married and lived with a spouse or partner.

Table 1.

Demographic Characteristics, Health, and Breast Cancer History (N = 190)

Variables	Mean	SD	range	n	%
Demographic characteristics
Age (years)	70.4	5.3	65 - 97
Education (years)	14	2.6	0 - 23
Less than high school				2	1.1
High school graduate				76	40.0
Vocational technical school				39	20.5
College				38	20.0
Graduate and over				34	17.9
Missing				1	0.5
Income ($)
< 10,000				7	3.7
10,000-24,999				40	21.1
25,000-39,999				39	20.5
40,000 or more				84	44.2
Missing				20	10.5
Ethnicity
Caucasian				178	93.6
African-American and others				12	6.4
Marital status
Married				111	58.5
Widowed				47	24.7
Divorced or separated				20	10.5
Never married				12	6.3
Living arrangement
Live alone				64	33.7
Live with someone else				126	66.3
Health status
Number of health problems	5.9	2.5	2 - 14
Number of symptoms	17.8	7.5	3 - 37
Number of medications	8.5	3.7	0 - 21
Breast cancer history
Months since cancer diagnosis	39.8	40.9	12 - 266
Months since last treatments	34.8	40.9	5 - 266
Breast cancer treatments
Radiation				135	71.1
Lumpectomy				129	67.9
Hormonal therapy				129	67.9
Mastectomy				75	39.5
Chemotherapy				72	37.9

Participants reported an average of six health problems other than cancer and 18 symptoms. The most frequently reported health problems were arthritis (74.7%), high cholesterol (59.5%) and high blood pressure (57.9%). The most frequently reported symptoms were itching (94.2%), stiffness (87.4%), pain (79.5%), aching (79.5%), and fatigue (77.9%).

Participants had been diagnosed with breast cancer an average of 40 months, and 86.3% of the participants had been diagnosed less than 5 years. The majority of the participants had undergone lumpectomy (n = 129, 67.9%) and radiation therapy (n = 135, 71.1%), and over half of the participants (n = 129, 67.9%) were currently receiving hormonal therapy. Average months since last treatment (not including hormonal) was 34.8 months.

Description of Beliefs and Quality of Life

Descriptive statistics for the SMBQ, CommA, and CommD scales indicated that participants had low levels of negative beliefs. The mean score of SMBQ was 1.55 (SD = 0.73, range = 0.17 – 4.46). The three items with the highest mean agreement were: “Many symptoms are just a normal part of getting older” (M = 3.75, SD =1.12), “It is easier to put up with pain than with the side effects of some medications” (M = 2.56, SD =1.45), and “The ‘cure’ for symptoms is often worse than the disease” (M = 2.03, SD =1.49).

Table 2 shows frequencies and percentages of each item on the CommA and CommD scales. The mean CommA score was 1.61 (SD = 2.06, range = 0 – 10), indicating that participants had perceived few negative attitudes from their HCPs. However, 60 percent (n = 114) of the participants reported at least one negative attitude on the CommA, and approximately 26 percent (n = 49) reported four or more negative attitudes. The most frequently endorsed item was “You’re worrying too much” (38.9 %), followed by “Compared to other women with breast cancer, your problems are minor” “It’s not necessary for you to know the details about your condition”, and “You should be thankful you’ve lived as long as you have.”

Table 2.

Frequency and Percentage of Older Breast Cancer Survivors Endorsing Items on CommA and CommD (N = 190)

Items	n	%
CommA
Being told, “You’re worrying too much.”	74	38.9
Being told, “Compared to other women with breast cancer, your  problems are minor.”	38	20.0
Being told, “It’s not necessary for you to know the details about your  condition.”	37	19.5
Hearing, “You should be thankful you’ve lived as long as you have.”	35	18.4
Feeling a symptom you reported was not dealt with to your satisfaction.	30	15.8
Being told, “You shouldn’t be so concerned, worried, or anxious about  breast cancer now.”	26	13.7
Being told that, “You’ll have to learn to live with it because your  symptom(s) are part of getting older.”	24	12.6
Being told, “A lot women with breast cancer are worse off than you.”	23	12.1
Being told, “Not much can be done about one or more of your  symptoms because they’re part of getting older.”	11	5.8
Hearing, “You’re better off than a lot of women who have had breast  cancer.”	8	4.2
CommD
Not knowing which doctor I should talk to about different symptoms.	79	41.6
Not knowing for sure which symptoms to tell my doctor about.	72	37.9
Not knowing if my doctors are communicating with each other  about my health care.	66	34.7
Not feeling comfortable talking about my symptoms because they may  seem minor, and I don’t want to be a “complainer.	52	27.4
Worry that I am ignoring symptoms that may be serious.	50	26.3
Worry that I will be labeled a “problem patient” or “neurotic”  if I talk about my symptoms or concerns.	19	10.0

Note. CommA = Communication Attitudes, CommD = Communication Difficulties

The mean CommD score was 1.78 (SD = 1.90, range = 0 – 6), indicating that participants had perceived few difficulties in communicating with HCPs about their symptoms. However, over 60 percent of participants (n = 116) reported at least one difficulty on the CommD, and 23.2 percent of participants (n = 44) reported four or more difficulties. In addition, 4.2 percent of participants (n = 8) reported all six communication difficulties. The most frequently endorsed item (41.6 %) was, “Not knowing which doctor I should talk to about different symptoms.” Over one-third of respondents reported, “Not knowing for sure which symptoms to tell my doctor about” and “Not knowing if my doctors are communicating with each other about my health care” as difficulties.

For the measures of quality of life, the mean standardized score on the PCS (M = 41.3, SD = 10.3, range = 16.3 – 59.7) was lower than the MCS (M = 53.6, SD = 9.1, range = 22.0 – 69.1). The PCS score of the sample was slightly lower than the population norm for adults in the U.S., aged 65 – 74 (M = 43.7, SD = 11.0, range = 13 – 59) (Ware et al., 2000). However, the MCS score in this study was slightly higher than the population norm for adults in the U.S., aged 65 – 74 (M = 52.1, SD = 9.53, range = 19 – 70) (Ware et al., 2000). The means for the PIL (M = 4.88, SD = 0.81, range = 1.86 – 6.0) and PR scales (M = 5.13, SD = 0.79, range = 2.93 – 6.0) indicate high levels of psychological well-being. These four scores reflect overall high quality of life of the participants, as expected in this community-dwelling sample (Robb, Haley, Balducci, Extermann, Perkins, & Small et al., 2007).

Relationships among Three Beliefs as Barriers to Symptom Management and Quality of Life

Bivariate correlations indicated that three beliefs were significantly, moderately, and positively related to each other: SMBQ and CommA (r = .40, p < .01), SMBQ and CommD (r = .44, p < .01) and CommA and CommD (r = .47, p < .01). Neither demographic characteristics nor breast cancer history were significantly correlated with any of the three beliefs. However, all three beliefs had significant and positive correlations with number of health problems (r’s ranged from .15 - .24, p < .05) and number of symptoms (r’s ranged from .27 - .36, p < .01).

Hypothesis 1 was tested by examining the direct effects of SMBQ and CommA on CommD after controlling for age, number of health problems, and number of symptoms. The direct paths for SMBQ (β = .86, b = 1.14, SE = .05 p < .000) and for CommA (β = .16, b = .20, SE = .08, p = .012) were significant, supporting hypothesis 1. Negative beliefs about symptom management and perceived negative attitudes from HCPs were significant predictors of more communication difficulties.

Hypothesis 2 was tested by examining the indirect effects of SMBQ and COMMA on each quality of life outcome separately, after controlling for age, number of health problems, and number of symptoms (See Table 3). Before testing the indirect effects, the direct effects of CommD on each quality of life outcome were tested. CommD had significant direct, negative effects on all four dimensions of quality of life (i.e., PCS β = − .15, b= −.60, SE= .30 p < .05; MCS β = − .20, b= −.73, SE= .26, p < .01; PIL β = − .26, b= −.08, SE= .02, p < .01; and PR β = −.28, b= −.09, SE= .03, p < .01).

Table 3.

Summary of Total Effects of SMBQ and CommA and Indirect Effects of SMBQ and CommA Through CommD on Four Dimensions of QOL (N = 190)

	PCS			MCS			PIL			PR
	B	SE	β	B	SE	β	B	SE	β	B	SE	β
SMBQ	−0.031	0.805	−0.006	−1.122	0.686	−0.234	−0.231**	0.063	−0.549**	−0.221**	0.065	−0.521**
CommD	−0.578	0.620	−0.148	0.027	0.528	0.008	0.073	0.048	0.228	0.059	0.050	0.184
Covariates
Age	−0.216	0.126	−0.111	−0.196	0.108	−0.109	−0.028**	0.010	−0.177**	−0.020*	0.010	−0.124*
Number of health problems	−1.209**	0.302	−0.298**	−0.138	0.257	−0.037	−0.004	0.024	−0.013	0.036	0.024	0.109
Number of symptoms	−0.327**	0.104	−0.265**	−0.558**	0.089	−0.487**	−0.038**	0.008	−0.381**	−0.037**	0.008	−0.362**
Total effect
SMB via CommD	−0.688	0.387	−0.133	−1.091**	0.329	−0.227**	−0.148**	0.030	−0.352**	−0.154**	0.031	−0.363**
Indirect effect
SMB via CommD	−0.658	0.706	−0.127	−0.031	0.602	−0.233	0.083	0.055	0.197	0.067	0.057	0.158
CommA	−0.010	0.338	−0.002	0.427	0.288	0.093	0.002	0.027	0.006	−0.009	0.028	−0.022
CommD	−0.579*	0.303	−0.153*	−0.800**	0.258	−0.220**	−0.084**	0.024	−0.263**	−0.089**	0.025	−0.278**
Covariates
Age	−0.216	0.127	−0.111	−0.220*	0.108	−0.122*	−0.030**	0.010	−0.192**	−0.022*	0.011	−0.136*
Number of health problems	−1.208**	0.304	−0.297**	−0.197	0.260	−0.052	−0.006	0.025	−0.018	0.036	0.025	0.109
Number of symptoms	−0.326**	0.105	−0.265**	−0.585**	0.089	−0.511**	−0.041**	0.008	−0.404**	−0.038**	0.009	−0.381**
Total effect
CommA via CommD	−0.130	0.335	−0.026	0.265	0.290	0.058	−0.027	0.029	−0.036	−0.027	0.029	−0.067
Indirect effect
CommA via CommD	−0.120	0.077	−0.024	−0.161*	0.602	−0.035*	−0.017*	0.008	−0.042*	−0.018*	0.009	−0.044*

Note. SMBQ = Symptom Management Beliefs Questionnaire, CommA = Communication Attitudes, CommD = Communication Difficulties PCS = Physical Component Summary from SF-12, MCS = Mental Component Summary from SF-12, PIL = Purpose in Life, PR = Personal Relations

^**p < .01,

^*p < .05

Mediating effects are shown when the indirect effects of SMBQ or CommA through CommD are significant. Hypothesis 2 was partially supported, as shown in Table 3. For SMBQ, total effects of SMBQ on MCS (β = −.23, p = .001), PIL (β = −.35, p < .001), and PR (β = −.36, p < .001) but not PCS, were statistically significant, indicating that higher levels of SMBQ predicted lower levels of quality of life in three of four dimensions. However, none of the four indirect effects of SMBQ on quality of life through CommD was significant, indicating that CommD does not mediate the effects of SMBQ on quality of life.

For CommA, the total effects of CommA on four quality of life measures were not significant. However, the indirect effects for MCS (β = −.035, p = .050), PIL (β = −.042, p < .05), and PR (β = −.044, p < .05) through CommD were significant (see Figure 2), indicating that CommD mediates the effects of CommA on MCS, PIL, and PR. That is, the negative effects of ageist attitudes from HCPs on quality of life can be explained in part by their detrimental effects on older women’s communication with their HCPs about symptom management.

Figure 2.

Direct and indirect effects of CommA on quality of life through CommD

Discussion

This study found that three beliefs that can be barriers to symptom management were associated with quality of life in older breast cancer survivors. In addition, the effect of perceived negative attitudes from HCPs on quality of life (i.e., MCS, PIL, and PR) was mediated by the difficulties cancer survivors experience in communicating with HCPs about symptoms and symptom management. As expected, perceiving negative attitudes of HCPs was associated with higher levels of communication difficulties with HCPs about symptoms, which, in turn, decreased quality of life. Further, negative beliefs about symptom management were related to lower quality of life regardless of the level of perceived communication difficulties.

There are a number of possible explanations for some of these findings. Older breast cancer survivors who hold ageist self-stereotypes about experiencing and managing symptoms may be more likely to perceive negative attitudes of HCPs in medical encounters. Or, perceiving more negative attitudes of HCPs may instill or reinforce negative beliefs about symptom management. In either case, it suggests that these mechanisms may be important to address to improve both patient-provider interactions and self-care on the part of older cancer survivors.

This study found that more negative symptom management beliefs and HCP attitudes were related to higher perceived communication difficulties. This finding is particularly troubling because the perception of communication difficulties may serve as a barrier to seeking health care for symptom management or engaging in self- care behaviors in older cancer survivors, which may subsequently influence health outcomes and reduce quality of life.

The most frequently endorsed items from each of the three beliefs scales are notable: “Many symptoms are just a normal part of getting older” (SMBQ); “You’re worrying too much” (CommA); and “Not knowing which doctor I should talk to about different symptoms” (CommD). These results are consistent with a previous study (Yeom & Heidrich, 2009). The first reflects ageist self-stereotypes that may impede seeking care for symptoms that they view as age-related or to discount the value of treatment. The second item implies that, although perhaps well-meaning, some statements by HCPs can be perceived dismissive of their real concerns. It may also reflect “beneficent ageism” on the part of HCPs. Beneficent ageism arises from misconceptions about aging and results in trivializing older patients’ desires, capacities, and goals for treatment but is meant with positive intent (Kagan, 2008). The third item indicates that, although older breast cancer survivors have multiple primary and specialty HCPs due to their numerous comorbidities, they have little guidance as to how to negotiate care. Experiencing symptoms may be worrisome because symptoms may or may not be cancer-related, and the ambiguity about the meaning of symptoms leads to uncertainty about the appropriate self care behavior. The importance of ambiguity about symptoms in old age deserves further attention, especially as it relates to improving self-care for symptom management.

Limitations

A number of limitations should be noted. Although recruitment was state-wide and targeted toward enhancing participation of racial and ethnic minority women, participants were primarily Caucasian, limiting the generalizability of the findings. Further, older women in this study may have been better off in terms of education and income than many older breast cancer survivors and some received their care in an academic medical center. Beliefs as barriers to symptom management might be different in those less well off or in different health care settings.

This study investigated beliefs as barriers to symptom management but not actual behaviors for symptom management. The relations with actual self-care and effective symptom management need to be investigated in future studies.

Certainly, differences in age-related beliefs about symptom management might exist for different genders, racial/ethnic groups, or older people with other health problems. In a prior comparison study between US and Japan, cultural differences related to beliefs about health in old age have been found (Levy et al., 2009). Further research should address these possible differences.

Lastly, this study was cross-sectional. As such, the interpretations of findings, particularly in terms of causality, are limited. Although it is postulated that the three beliefs lower quality of life, it is also possible that the reverse effect is true. A lower quality of life may aggravate detrimental perceptions and attitudes about symptom management. These hypotheses would be important to address in future longitudinal research.

Conclusions and Implications

This study found significant relationships between specific age-related beliefs about symptoms and symptom management on quality of life in older breast cancer survivors. The findings provide insights into future psycho-educational interventions that could be developed and tested to improve self-care for symptom management of older cancer survivors. Prior research has shown that older people respond with enhanced adherence and better health outcomes when interventions provide strategies that participants could control along with psychological support (Gitlin, Winter, Dennis, Corcoran, Schinfeld, & Hauck, 2006). The results of this study also suggest the importance of providing opportunities for older people to express their own views and emotions about aging. In addition, addressing stereotyped and erroneous beliefs about aging and their detrimental psychosocial consequence may be crucial for older individuals to explore their own misconceptions in experiencing and managing symptoms and thereby to guide them in developing beneficial self-care strategies.

On the other hand, the significant mediating role of CommD on the relationship between CommA and quality of life provides insights into the critical role of patient-provider communication. Health care providers should be cognizant of stereotyped beliefs about aging that may affect self-care, management, and health outcomes of older adults. Patients and providers may have different health care goals. For instance, in one study, HCPs tended to focus more on strategies to increase survival length, while older cancer survivors were eager to receive information not only about the illness but also about their overall health and daily life (Spagnola, Zabora, BrintzenhofeSzoc, Hooker, Cohen, & Baker, 2003). Such discordant expectations may lead older patients to feel that their concerns are being ignored (Nussbaum et al., 2003; Spagnola et al., 2003). Ultimately, HCPs need to be introspective about their own beliefs and attitudes toward older patients and their health problems. Both older patients and HCPs must attempt to overcome stigmatized beliefs about aging in experiencing and managing symptoms.