Abstract
BACKGROUND
Given differences in cancer survival by race, black women may differ from white women in breast cancer risk perceptions.
OBJECTIVE
To evaluate black-white differences in risk perceptions of breast cancer survival and screening mammography benefit.
DESIGN
A written survey was administered to a random sample of women attending general internal medicine clinics.
PARTICIPANTS
Black and white women, ages 40 to 69.
MEASUREMENTS
Risk perceptions were measured regarding (1) average 5-year survival after a breast cancer diagnosis and (2) relative risk reduction of screening mammography. Women's risk perceptions were defined as being accurate, as well as more or less pessimistic. Measured patient characteristics included race, age, family history of breast cancer, income, insurance, education, and numeracy. Unadjusted Pearson χ2 tests and adjusted multivariable regression analyses were done.
RESULTS
Black women were more likely than white women to accurately perceive breast cancer survival in both unadjusted (48% vs 26%, P <.001) and adjusted analyses (adjusted odds ratio (AOR)=3.58; 95% confidence interval (CI)=1.56 to 8.21). Black women were also more likely to accurately perceive the benefit of screening mammography in unadjusted (39% vs 15%, P <.001) and adjusted analyses (AOR=2.70; 95% CI=1.09 to 6.69). Black women were more likely to have a more pessimistic perception of mammography benefit in unadjusted (47% vs 15%, P <.0001) and adjusted analyses (AOR=3.94; 95% CI=1.62 to 9.56).
CONCLUSIONS
Awareness of risk perceptions can help physicians to tailor patient education. Physician acknowledgment of more accurate risk perceptions among black women may serve as a basis to improve patient-physician communication.
Keywords: breast cancer, cancer screening, race & ethnicity, risk assessment
Women's perceptions of breast cancer risk provide valuable insights to physicians who are trying to communicate effectively with their patients. The accuracy of risk perceptions among patients is one measure of the effectiveness of patient-physician communication. The interpretation of facts may be affected by powerful emotions given the charged nature of cancer as a subject, but those exposed to health risks should have some understanding about the size of the threat.1 Accurate risk perception is important because it is a necessary condition for informed decision making among patients. Potential barriers to the successful communication of risk include ambiguous or inaccurate messages transmitted by the physician, as well as the patient's inability to understand quantitative health information.
Patient factors that influence risk perception have been studied previously. Older age was associated with the overestimation of breast cancer risk prior to genetic counseling in a study of high-risk patients in Scotland.2 In a study of patients from New England, women with less education and less numeracy appeared to have less accurate perceptions of risk.3 Yet information regarding the association of race and the perception of breast cancer risks is incomplete. There is empirical evidence to suggest that black patients perceive medical risks differently than white patients. For example, in the case of end-stage renal disease, black men were less likely to expect longer survival with renal transplant versus dialysis,4 despite evidence in the medical literature to the contrary.5
Previous research has described the presence of fatalism among black patients,6–8 or the belief that events are predetermined and individuals are powerless to change them. Previous research has also described mistrust of the health care system among black individuals,9 or a feeling that their medical needs will not necessarily be put above all other considerations. However, less is known about whether black women have different perceptions of quantitative risks. In this study we evaluated perceptions of (a) breast cancer survival and (b) the benefit of screening mammography among a group of black and white survey participants. We tested whether black women have more or less accurate perceptions of the risks associated with breast cancer than white women. We also tested whether black women have more or less pessimistic perceptions of breast cancer risks than white women. Pessimism was defined as the perception of a higher quantitative risk of experiencing an adverse outcome.
METHODS
Survey Protocol and Population
We surveyed a random sample of women who attended two general internal medicine clinics at an academic medical center in Milwaukee, Wisconsin. Data used for this study was part of the baseline evaluation of a clinical trial evaluating the effect of tailored risk information on breast cancer risk perceptions. Inclusion criteria for the trial included female gender, age 40 to 85, and English-speaking. Exclusion criteria included a personal history of breast cancer, dementia, or a life expectancy of less than 2 years as judged by the primary care physician.
Women who were sampled were sent a recruitment letter that described a study about preventive health but did not refer specifically to breast cancer. Recruitment letters were sent between June 15, 1999, and June 19, 2000. The recruitment letters were followed up with a telephone call to the patient. After obtaining informed consent, the survey was self-administered by the patient in the clinic, with a research assistant available to answer questions about how to complete the survey. Enrolled subjects were compensated $20 for their time. Approval from the Human Subjects Review Committee for the participating institutions was obtained.
Recruitment letters were mailed to 1,409 women. Of the 1,409 eligible subjects, 967 were successfully contacted by telephone, and 254 women were enrolled and completed the survey. The final participation rate was 18% (254/1,409). Over 80% of those that declined stated the reason as lack of interest in participating in the study.
The following groups who were recruited for the initial clinical trial were excluded from the current study. Four Latino and two American Indian or Alaskan Native women were excluded because the primary research question of the current study addressed black-white differences in risk perception; in addition, there were inadequate numbers from these other groups to make meaningful comparisons. Forty-one women ages 70 to 84 were also excluded because of inadequate data from clinical trials to assess the benefit of screening mammography among this age group according to the U.S. Preventive Services Task Force at the time of this study.10 The final cohort for analysis included 207 black or white women ages 40 to 69.
Survey Instrument
Sociodemographic and clinical characteristics assessed as predictors of risk perception included the following: patient race, age, family history of breast cancer (≥1 first-degree relative), family income, health insurance, education, and numeracy. Race was measured by self-report using the following categories: “White or Caucasian, but not Hispanic or Latino” or “Black or African American, but not Hispanic or Latino.”
Numeracy was measured with a 3-item instrument adapted from Schwartz et al.11 designed to measure a patient's facility with basic probability and numerical concepts. Numeracy values based on the instrument range from 0 to 3, reflecting the number of correct responses to three questions. Patient numeracy was dichotomized into two categories for our models: numerate (three out of three correct responses) or innumerate (0 to 2 correct responses).
The primary outcomes measured were risk perceptions of (1) breast cancer survival and (2) screening mammography benefit. Risk perceptions of breast cancer survival were measured with the following survey item: “On average, when women get breast cancer what are their chances of living for 5 years or longer?” Risk perceptions of screening mammography benefit were measured with the following item: “For women your age, how much do you think regular mammograms decrease the risk of dying from breast cancer?” Responses were collected using a relative risk reduction format and a close-ended interval scale. The response intervals were chosen to detect large differences in risk perception. The response scale included the following response options: 0% to 25%, 26% to 50%, 51% to 75%, and 76% to 100%. For the screening mammography item, the lowest response category was further divided into “not at all” and “between 5% and 25%.” The definitions of accuracy and pessimism are described below.
Accuracy
The definition of an accurate response to the breast cancer survival item was race-specific and based upon data available at the time of the study from the Surveillance, Epidemiology, and End Results Program (1989–1996)12 Using this data source, an accurate perception of 5-year survival was defined as 71% among black women and 86% among white women.
The definition of an accurate response to the question regarding screening mammography benefit was age-specific and based upon a previous meta-analysis combining multiple studies of screening mammography.13 Screening mammography was considered to reduce the relative risk of dying of breast cancer by 26% (95% CI, 17% to 34%) among women ages 50 to 69 years, and by 7% (95% CI, −13% to 24%) among women ages 40 to 49 years. By accepting responses falling within the 95% confidence intervals described here as accurate, an accurate perception of screening benefit was defined as answering either “5% to 25%” or “26% to 50%” among women ages 50 to 69 years, and “not at all” or “5% to 25%” among women ages 40 to 49 years.
Pessimism
Pessimism was defined as a dichotomous variable in terms of risk perception. For the question regarding breast cancer survival, a more pessimistic perception would be to answer that there was somewhere between a 0% and 50% chance of living 5 years or longer when women get breast cancer, while a less pessimistic perception would be to answer between 51% and 100%. Similarly, for the question regarding screening mammography benefit, a more pessimistic perception would be to answer that regular mammograms decrease the risk of dying from breast cancer 0% to 50%, while a less pessimistic perception would be 51% to 100%.
Statistical Analysis
Pearson χ2 tests of significance were performed to evaluate the association between race and potential covariates, as well as the association between race and the outcomes of accurate and pessimistic risk perception. Logistic regression models were used to determine unadjusted odds ratios (ORs) as a measure of the association between race and the risk perception outcomes, as well as the association between potential covariates and the outcomes. Multivariable logistic regression models were constructed to predict the association of race with accurate and pessimistic risk perception while controlling for potential clinical (age (40 to 49 vs 50 to 69) and family history of breast cancer) and non-clinical covariates (income (<$20,000/year vs ≥$20,000/year), insurance type, level of education, and numeracy). All analyses were conducted using Stata version 8.0 (Stata Corp., College Station, TX).
Conflict of Interest Notification
This work was supported by a grant from the American Cancer Society. The funding agency approved the original research design in the form of a grant application. The funding agency had no further role in the research.
RESULTS
Study Population
Of the 207 persons in the cohort, 31% were black and 69% were white. The mean age was 55. Overall, black women were younger than white women (Table 1). There was no difference in family history of breast cancer by race. Black women were more likely than white women to have a family income of <$20,000 (80% vs 35%, P <.001), public insurance (77% vs 41%, P <.001), and to have not graduated from high school (33% vs 8%, P <.001).
Table 1.
Patient Characteristics by Race
| Patient Characteristics | Race | ||
|---|---|---|---|
| Black Women(N=64)n (%) | White Women(N=143)n (%) | P Value* | |
| Age | =.01 | ||
| 40 to 49 | 26 (41) | 34 (24) | |
| 50 to 69 | 38 (59) | 109 (76) | |
| ≥1 first-degree relative with breast cancer | 11 (17) | 27 (19) | =.77 |
| Insurance | <.001 | ||
| Private fee-for-service | 4 (6) | 60 (42) | |
| Health maintenance organization | 3 (5) | 17 (12) | |
| Medicare | 22 (34) | 33 (23) | |
| Medicaid or Milwaukee County | 28 (44) | 26 (18) | |
| General Assistance Program | |||
| None or other | 7 (11) | 7 (5) | |
| Family income | <.001 | ||
| <$20,000 | 51 (80) | 50 (35) | |
| ≥$20,000 | 13 (20) | 93 (65) | |
| Education | <.001 | ||
| Less than high school | 21 (33) | 12 (8) | |
| High school graduate | 39 (61) | 89 (62) | |
| College graduate | 4 (6) | 19 (13) | |
| Post-graduate | 0 (0) | 23 (16) | |
Pearson χ2 test of association.
Breast Cancer Survival
Among women overall, 33% accurately perceived the average chances of living 5 years or longer after a breast cancer diagnosis, 4% overestimated the chances, and 63% underestimated the chances. A greater proportion of black women accurately perceived breast cancer survival than white women (48% vs 26%, P <.001). A greater proportion of black women were also more pessimistic than white women (38% vs 22%, P =.02) regarding breast cancer survival (Table 2).
Table 2.
Perceptions of Breast Cancer Survival by Race
| Patient Race | Chances of Living 5 y or Longer | |||
|---|---|---|---|---|
| More Pessimistic | Less Pessimistic | |||
| 0% to 25% | 26% to 50% | 51% to 75% | 76% to 100% | |
| Black (n=64) | 8 (13) | 16 (25) | 31 (48)* | 9 (14) |
| White (n=143) | 6 (4) | 25 (18) | 75 (52) | 37 (26)* |
| All women | 14 (7) | 41 (20) | 106 (51) | 46 (22) |
Accurate response to question regarding breast cancer survival: “On average, when women get breast cancer what are their chances of living for 5 y or longer?” An accurate perception of 5-year survival was defined as 71% among black women and 86% among white women.
In multivariable analysis, black women remained more likely than white women to accurately perceive 5-year breast cancer survival (AOR=3.58; 95% CI=1.56 to 8.21), (Table 3). No other predictor besides race was associated with more accurate perceptions of breast cancer survival in multivariable analysis. Black women were no more likely than white women to have a more pessimistic perception of breast cancer survival in multivariable analysis (AOR=1.49; 95% CI=0.67 to 3.32).
Table 3.
Perceptions of Breast Cancer Survival by Race and Other Patient Characteristics
| Patient Characteristic | Accurate | More Pessimistic | ||||||
|---|---|---|---|---|---|---|---|---|
| OR* (95% CI) | P Value | AOR† (95% CI) | P Value | OR* (95% CI) | P Value | AOR† (95% CI) | P Value | |
| Race | ||||||||
| White | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| Black | 2.69 (1.45 to 4.99) | .002 | 3.58 (1.56 to 8.21) | .003 | 2.17 (1.14 to 4.13) | .02 | 1.49 (0.67 to 3.32) | .33 |
| Age | ||||||||
| 40 to 49 | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| 50 to 69 | 0.47 (0.25 to 0.88) | .02 | 0.60 (0.30 to 1.21) | .15 | 1.65 (0.80 to 3.42) | .17 | 1.78 (0.80 to 3.98) | .16 |
| Family history | ||||||||
| None | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| ≥1 relative | 1.43 (0.69 to 2.95) | .34 | 1.52 (0.69 to 3.34) | .30 | 1.35 (0.63 to 2.91) | .44 | 1.18 (0.52 to 2.68) | .69 |
| Income | ||||||||
| <$20,000/yr | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| ≥$20,000/yr | 1.21 (0.68 to 2.17) | .52 | 2.47 (0.97 to 6.30) | .06 | 0.44 (0.23 to 0.83) | .01 | 0.78 (0.31 to 1.96) | .60 |
| Insurance | ||||||||
| Private FFS | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| HMO | 0.64 (0.20 to 1.98) | .44 | 0.63 (0.19 to 2.06) | .44 | 0.21 (0.03 to 1.69) | .14 | 0.14 (0.02 to 1.25) | .08 |
| Medicare | 0.72 (0.33 to 1.57) | .41 | 0.81 (0.29 to 2.24) | .68 | 2.42 (1.07 to 5.48) | .03 | 1.12 (0.40 to 3.14) | .84 |
| Public ins | 1.21 (0.57 to 2.58) | .61 | 1.32 (0.44 to 3.90) | .62 | 1.65 (0.71 to 3.84) | .24 | 0.76 (0.24 to 2.38) | .64 |
| None/other | 1.06 (0.32 to 3.55) | .92 | 0.78 (0.18 to 3.30) | .73 | 1.57 (0.42 to 5.81) | .50 | 0.78 (0.18 to 3.35) | .74 |
| Education | ||||||||
| <High school | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| High school | 0.98 (0.43 to 2.20) | .96 | 1.16 (0.45 to 3.02) | .76 | 0.38 (0.17 to 0.85) | .02 | 0.64 (0.26 to 1.57) | .33 |
| College | 1.29 (0.43 to 3.89) | .66 | 1.57 (0.42 to 5.93) | .50 | 0.42 (0.13 to 1.35) | .15 | 0.82 (0.22 to 3.10) | .77 |
| Post–grad | 0.71 (0.22 to 2.29) | .56 | 1.05 (0.25 to 4.37) | .95 | 0.18 (0.04 to 0.73) | .02 | 0.38 (0.08 to 1.94) | .25 |
| Numeracy | ||||||||
| Innumerate | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| Numerate | 0.73 (0.40 to 1.31) | .29 | 0.84 (0.38 to 1.85) | .67 | 0.43 (0.22 to 0.84) | .01 | 0.60 (0.26 to 1.38) | .23 |
Odds ratio (OR) from logistic regression model using single patient characteristic as predictor.
Adjusted odds ratio (AOR) from multivariable logistic regression model using patient race, age, family history, family income, insurance, education, and numeracy as predictors.
FFS, fee-for-service; HMO, health maintenance organization; yr, year; ins, insurance; Post-grad, Post-graduate.
Screening Mammography Benefit
Among women overall, 23% accurately perceived how much regular screening mammograms decrease the risk of dying from breast cancer, 1% of women underestimated mammography benefit, and 76% overestimated mammography benefit. A greater proportion of black women accurately perceived mammography benefit than white women (39% vs 15%, P <.001). A greater proportion of black women were also more pessimistic than white women (47% vs 15%, P <.001) about mammography benefit (Table 4).
Table 4.
Perceptions of Screening Mammography Benefit by Race and Age
| Patient Age and Race | Mammograms Decrease Risk of Dying of Breast Cancer | ||||
|---|---|---|---|---|---|
| More Pessimistic | Less Pessimistic | ||||
| Not at all | 5% to 25% | 26% to 50% | 51% to 75% | 76% to 100% | |
| Age 40 to 49 | |||||
| Black (n=26) | 3 (12)* | 5 (19)* | 3 (12) | 5 (19) | 10 (38) |
| White (n=34) | 0* | 2 (6)* | 0 | 13 (38) | 19 (56) |
| All women, age 40 to 49 | 3 (5)* | 7 (12)* | 3 (5) | 18 (30) | 29 (48) |
| Age 50 to 69 | |||||
| Black (n=38) | 2 (5) | 7 (18)* | 10 (26)* | 7 (18) | 12 (32) |
| White (n=109) | 0 | 6 (6)* | 14 (13)* | 37 (34) | 52 (48) |
| All women, age 50 to 69 | 2 (1) | 13 (9)* | 24 (16)* | 44 (30) | 64 (44) |
Accurate response to question regarding screening mammography benefit: “For women your age, how much do you think regular mammograms decrease the risk of dying from breast cancer?” The accurate perception of screening mammography benefit was considered to be a relative risk reduction of 26% (95% CI: 17% to 34%) among women age 50 to 69 y, and of 7% (95% CI: −13% to 24%) among women ages 40 to 49 y. Responses within the 95% confidence intervals described here were considered to be accurate.
In multivariable analysis, black women were more likely than white women to accurately perceive mammography benefit (AOR=2.70; 95% CI=1.09 to 6.69), (Table 5). Black women were also more likely to have a pessimistic perception of mammography benefit in multivariable analysis (AOR=3.94; 95% CI=1.62 to 9.56). Also in multivariable analysis, women who graduated from high school were less likely than women who did not graduate from high school to accurately perceive mammography benefit (AOR=0.16; 95% CI=0.06 to 0.44) and less likely to have a pessimistic perception of mammography benefit (AOR=0.18; 95% CI=0.07 to 0.49).
Table 5.
Perceptions of Screening Mammography Benefit by Race and Other Patient Characteristics
| Patient Characteristic | Accurate | More Pessimistic | ||||||
|---|---|---|---|---|---|---|---|---|
| OR* (95% CI) | P Value | AOR† (95% CI) | P Value | OR* (95% CI) | P Value | AOR† (95% CI) | P Value | |
| Race | ||||||||
| White | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| Black | 3.53 (1.79 to 6.94) | <.001 | 2.70 (1.09 to 6.69) | .03 | 4.85 (2.49 to 9.47) | <.001 | 3.94 (1.62 to 9.56) | .002 |
| Age | ||||||||
| 40 to 49 | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| 50 to 69 | 1.68 (0.78 to 3.65) | .19 | 2.05 (0.82 to 5.12) | .13 | 1.31 (0.64 to 2.67) | .47 | 1.59 (0.67 to 3.79) | .29 |
| Family history | ||||||||
| None | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| ≥1 relative | 1.27 (0.58 to 2.86) | .56 | 1.24 (0.49 to 3.13) | .65 | 1.27 (0.58 to 2.79) | .55 | 1.32 (0.53 to 3.28) | .55 |
| Income | ||||||||
| <$20,000/yr | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| ≥$20,000/yr | 0.45 (0.23 to 0.89) | .02 | 0.82 (0.28 to 2.37) | .71 | 0.40 (0.21 to 0.78) | .006 | 0.76 (0.27 to 2.17) | .61 |
| Insurance | ||||||||
| Private FFS | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| HMO | 3.00 (0.89 to 10.06) | .08 | 2.61 (0.68 to 9.98) | .16 | 3.00 (0.89 to 10.06) | .08 | 2.76 (0.72 to 10.59) | .14 |
| Medicare | 1.95 (0.73 to 5.20) | .18 | 0.55 (0.14 to 2.11) | .38 | 2.39 (0.92 to 6.23) | .07 | 0.72 (0.19 to 2.66) | .62 |
| Public ins | 2.69 (1.04 to 6.96) | .04 | 0.80 (0.20 to 3.25) | .76 | 3.22 (1.26 to 8.21) | .02 | 0.92 (0.23 to 3.63) | .90 |
| None/other | 5.25 (1.44 to 19.11) | .01 | 3.00 (0.65 to 13.96) | .16 | 7.00 (1.94 to 25.25) | 3.93 (0.84–18.44) | .08 | |
| Education | ||||||||
| <High school | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| High school | 0.14 (0.06 to 0.33) | <.001 | 0.16 (0.06 to 0.44) | <.001 | 0.15 (0.07 to 0.34) | <.001 | 0.18 (0.07 to 0.49) | .001 |
| College | 0.11 (0.03 to 0.45) | .002 | 0.14 (0.03 to 0.74) | .02 | 0.14 (0.04 to 0.49) | .002 | 0.21 (0.04 to 1.02) | .05 |
| Post-grad | 0.16 (0.04 to 0.56) | .004 | 0.27 (0.05 to 1.36) | .11 | 0.14 (0.04 to 0.49) | .002 | 0.33 (0.07–1.68) | .18 |
| Numeracy | ||||||||
| Innumerate | 1.0 | 1.0 | 1.0 | 1.0 | ||||
| Numerate | 0.31 (0.15 to 0.66) | .002 | 0.75 (0.28 to 2.02) | .57 | 0.30 (0.14–0.61) | .001 | 0.86 (0.33 to 2.26) | .76 |
Odds ratio (OR) from logistic regression model using single patient characteristic as predictor.
Adjusted odds ratio (AOR) from multivariable logistic regression model using patient race, age, family history, family income, insurance, education, and numeracy as predictors.
FFS, fee-for-service; HMO, health maintenance organization; yr, year; ins, insurance; Post-grad, Post-graduate.
DISCUSSION
The majority of women overestimated the risk of dying of breast cancer and overestimated the benefit of screening mammography. These findings are similar to prior research demonstrating that women commonly overestimate their likelihood of developing or dying of breast cancer when asked to define their risk on an open-ended numeric scale.14–17 Women younger than 50 years of age have also been previously shown to overestimate the benefit of screening mammography.15
In the case of breast cancer survival, black women were more likely to have accurate perceptions than white women in adjusted analysis, as well as more pessimistic perceptions in an unadjusted comparison. More pessimistic perceptions among black women, in fact, tend to be more accurate when one considers that 5-year breast cancer survival is persistently lower among black than white patients.12,18 Black women may be more accurate simply because they have more knowledge or information about breast cancer survival. Black women may have a greater, heightened awareness of their 5-year survival rates because these rates are lower than among white women.
Greater fatalism among black women may also contribute to more pessimistic, and thus more accurate, perceptions of breast cancer survival.6–8 Lannin et al.6 assessed fatalism by asking whether black women believed several statements were true, including “if it's meant to be, I will stay healthy.” Based upon focus group interviews, Phillips et al.7 concluded that black women had “fatalistic perspectives regarding breast cancer outcomes”. In a literature review, Powe8 described multiple operational definitions of fatalism, and in her own approach, described cancer fatalism as having the following attributes: fear, pessimism, predetermination, and a sense of inevitable death. Fatalism of these types may contribute to the more pessimistic, and thus more accurate, perceptions of breast cancer survival observed among black women in our study.
In the case of screening mammography, black women were more likely than white women to have accurate, as well as more pessimistic, perceptions of the benefit of screening. There are several possible explanations for this finding. First, black women may have more knowledge or information about the benefit of screening mammography. Second, fatalism among black women may contribute to skepticism about the chances that screening mammography will change the course of disease, and thus, may lead to more pessimistic risk perceptions among this group of women. A group of women with more pessimistic perceptions of screening benefit will tend to be more accurate when found within a population whose perceptions of benefit are inflated overall (as was the case for the full population in our study). Third, mistrust among black women may also contribute to more guarded perceptions about the benefit of screening interventions recommended by physicians. Trust was found by Boulware et al.9 to be lower among black patients when operationally defined as a patient's agreement with the following statement: “I trust my physician will put my medical needs above all other considerations when treating my medical problems.”
Lower levels of education were associated with more accurate, as well as more pessimistic, perceptions of mammography benefit. The association between higher levels of education and less accurate perceptions of mammography benefit is intriguing. Perhaps the primary mechanism through which greater educational achievement influences risk perception is the creation of an optimistic set of expectations, as opposed to the transmission of critical thinking skills that are presumed necessary to generate accurate risk estimates. Such high expectations may result in the creation of less pessimistic, as opposed to more accurate, risk estimates.
It is possible that the local health care system and/or the physicians of patients in the community studied here communicate accurate information more effectively to black women. Black women in this study may also be drawn from communities with greater shared knowledge about breast cancer. Future research should explore how much the health care system or local communities contribute to individual perceptions of risk. To the extent that local factors influence the findings reported here, the results may not be generalizable to a national population. However, previous studies of breast cancer risk perception have included little information about black women,2,3,14 –16 and their inclusion is a strength of this study.
This study has limitations. First, the participation rate of 18% means that the findings may be influenced by selection or volunteer bias, and thus reduces the study's external validity. Commonly, volunteers are healthier or have fewer competing time demands than nonvolunteers. Health status or competing demands may be associated with risk perceptions and may differ between black and white women. We are not able to assess differences in response rates between black and white women because race data was not collected among non-participants. However, recruitment purposefully took place in a clinic setting that served significant proportions of black patients. Thirty percent of the women in this study's population were black, compared with 13% of the U.S. civilian population.19 The black women in this study also had lower levels of education and income than the U.S. black population19; although these socioeconomic factors were included in the final multivariable regression models, this distribution makes the results most generalizable to a disadvantaged black population.
Second, the definition of pessimism was based upon quantitative risk perceptions. The related constructs of fatalism and mistrust were not measured directly. However, the definition of pessimism used has face validity, as well as a potential relationship to the constructs of fatalism and mistrust as discussed above. Qualitative and quantitative research would be helpful to further elucidate the relationships among pessimism, fatalism, and mistrust.
Another limitation of this study involves the categories used for measurement. The response categories are wide enough that an individual estimating at the extremes of the range for an accurate response may be qualitatively different than an individual estimating closer to the point estimate. Nonetheless, using a generous definition of what constitutes an accurate response, the majority of women were inaccurate and significant black-white differences were observed.
Our findings suggest that different challenges in patient-physician communication are posed by different types of patients seen in clinical practice. The challenge in communication among women who overestimate risk or benefit may be to moderate inflated perceptions. The challenge among other women, in particular black women, may be to find positive ways to confirm accurate perceptions of more modest screening benefit and worse survival, while at the same time, to communicate why guidelines currently recommend regular screening mammography based upon the available evidence.20 More pessimistic perceptions of risk, even though more accurate, may not necessarily be welcome knowledge from the patient's perspective. Furthermore, it would be prudent to assume that discussions of breast cancer risks take place among women outside the medical setting. Open, informed discussions of risk should also take place between patients and physicians within the medical encounter.
In summary, this study finds that black women have more accurate risk perceptions than white women with respect to breast cancer survival and screening mammography benefit. Awareness of risk perceptions among special populations can help physicians to tailor patient education. Goals of communication should include improving the accuracy of risk perceptions overall, and in some groups, openly discussing possible reasons for more pessimistic risk perceptions, including fatalism and mistrust. Among black women, physician acknowledgement of more accurate risk perceptions may serve as a confidence-building measure to increase patient trust. By placing breast cancer risks in an appropriate context relative to other health risks,21 physicians may lessen the cancer fatalism previously described among black women. Clinicians should be ready to discuss more pessimistic risk perceptions when present, and to acknowledge when these perceptions are more accurate, in order to improve patient-physician communication.
Acknowledgments
Dr. Haggstrom was supported by the VA Ambulatory Care Fellowship and Health Resources and Services Administration Faculty Development Grant (1D14 HP001 78-01). Dr. Schapira was supported by the American Cancer Society Cancer Control Career Development Award.
REFERENCES
- 1.Fischhoff B. Why (cancer) risk communication can be hard. J Natl Cancer Inst Monogr. 1999;25:7–13. doi: 10.1093/oxfordjournals.jncimonographs.a024213. [DOI] [PubMed] [Google Scholar]
- 2.Cull A, Anderson ED, Campbell S, Mackay J, Smyth E, Steel M. The impact of genetic counselling about breast cancer risk on women's risk perceptions and levels of distress. Br J Cancer. 1999 Feb;79:501–8. doi: 10.1038/sj.bjc.6690078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Woloshin S, Schwartz LM, Black WC, Welch HG. Women's perceptions of breast cancer risk: how you ask matters. Med Decis Making. 1999;19:221–9. doi: 10.1177/0272989X9901900301. [DOI] [PubMed] [Google Scholar]
- 4.Ayanian JZ, Cleary PD, Weissman JS, Epstein AM. The effect of patients' preferences on racial differences in access to renal transplantation. N Engl J Med. 1999;341:1661–9. doi: 10.1056/NEJM199911253412206. [DOI] [PubMed] [Google Scholar]
- 5.Ojo AO, Port FK, Wolfe RA, et al. Comparative mortality risks of chronic dialysis and cadaveric transplantation in black end-stage renal disease patients. Am J Kidney Dis. 1994;24:59–64. doi: 10.1016/s0272-6386(12)80160-0. [DOI] [PubMed] [Google Scholar]
- 6.Lannin DR, Mathews HF, Mitchell J, Swanson MS, Swanson FH, Edwards MS. Influence of socioeconomic and cultural factors on racial differences in late-stage presentation of breast cancer. JAMA. 1998;279:1801–7. doi: 10.1001/jama.279.22.1801. [DOI] [PubMed] [Google Scholar]
- 7.Phillips JM, Cohen MZ, Moses G. Breast cancer screening and African American women: fear, fatalism, and silence. Oncol Nurs Forum. 1999;26:561–71. [PubMed] [Google Scholar]
- 8.Powe BD. Cancer fatalism among Blacks: a review of the literature. Nurs Outlook. 1996;44:18–21. doi: 10.1016/s0029-6554(96)80020-0. [DOI] [PubMed] [Google Scholar]
- 9.Boulware LE, Cooper LA, Ratner LE, LaVeist TA, Powe NR. Race and trust in the health care system. Public Health Rep. 2003;118:358–65. doi: 10.1016/S0033-3549(04)50262-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.U.S. Preventive Services Task Force. Screening for Breast Cancer. Rockville, MD: Agency for Healthcare Research and Quality; 1996. [January 10, 2005]. Available at: http://www.ahrq.gov/clinic/2ndcps/brstcan.pdf. [Google Scholar]
- 11.Schwartz LM, Woloshin S, Black WC, Welch HG. The role of numeracy in understanding the benefit of screening mammography. Ann Intern Med. 1997;127:966–72. doi: 10.7326/0003-4819-127-11-199712010-00003. [DOI] [PubMed] [Google Scholar]
- 12.Greenlee RT, Hill-Harmon MB, Murray T, Thun M. Cancer statistics, 2001. CA Cancer J Clin. 2001;51:15–36. doi: 10.3322/canjclin.51.1.15. [DOI] [PubMed] [Google Scholar]
- 13.Kerlikowske K, Grady D, Rubin SM, Sandrock C, Ernster VL. Efficacy of screening mammography. A meta-analysis. JAMA. 1995;273:149–54. [PubMed] [Google Scholar]
- 14.Smith BL, Gadd MA, Lawler C, et al. Perception of breast cancer risk among women in breast center and primary care settings: correlation with age and family history of breast cancer. Surgery. 1996;120:297–303. doi: 10.1016/s0039-6060(96)80301-1. [DOI] [PubMed] [Google Scholar]
- 15.Black WC, Nease RF, Jr, Tosteson AN. Perceptions of breast cancer risk and screening effectiveness in women younger than 50 years of age. J Natl Cancer Inst. 1995;87:720–31. doi: 10.1093/jnci/87.10.720. [DOI] [PubMed] [Google Scholar]
- 16.Lerman C, Lustbader E, Rimer B, et al. Effects of individualized breast cancer risk counseling: a randomized trial. J Natl Cancer Inst. 1995;87:286–92. doi: 10.1093/jnci/87.4.286. [DOI] [PubMed] [Google Scholar]
- 17.Davids SL, Schapira MM, McAuliffe TL, Nattinger AB. Predictors of pessimistic breast cancer risk perceptions in a primary care population. J Gen Intern Med. 2004;19:310–5. doi: 10.1111/j.1525-1497.2004.20801.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Jemal A, Murray T, Samuels A, Ghafoor A, Ward E, Thun MJ. Cancer statistics, 2003. CA Cancer J Clin. 2003;53:5–26. doi: 10.3322/canjclin.53.1.5. [DOI] [PubMed] [Google Scholar]
- 19.McKinnon Jesse. The Black Population in the United States: March. Washington, DC: U.S. Census Bureau Current Population Reports; 2002. [October 17, 2005]. Series P20-541. 2003. Available at: http://www.census.gov/prod/2003pubs/p20-541.pdf. [Google Scholar]
- 20.U.S. Preventive Services Task Force. Screening for Breast Cancer: Recommendations and Rationale. Rockville, MD: Agency for Healthcare Research and Quality; February 2002. [February 15, 2005]. Available at: http://www.ahrq.gov/clinic/3rduspstf/breastcancer/brcanrr.htm. [Google Scholar]
- 21.Woloshin S, Schwartz LM, Welch HG. Risk charts: putting cancer in context. J Natl Cancer Inst. 2002;94:799–804. doi: 10.1093/jnci/94.11.799. [DOI] [PubMed] [Google Scholar]