Fax: (404) 325-9341
Sociodemographic determinants of cancer treatment health literacy
Article first published online: 27 JUN 2005
Copyright © 2005 American Cancer Society
Volume 104, Issue 3, pages 653–660, 1 August 2005
How to Cite
Gansler, T., Henley, S. J., Stein, K., Nehl, E. J., Smigal, C. and Slaughter, E. (2005), Sociodemographic determinants of cancer treatment health literacy. Cancer, 104: 653–660. doi: 10.1002/cncr.21194
Fax: (404) 325-9341
- Issue published online: 18 JUL 2005
- Article first published online: 27 JUN 2005
- Manuscript Accepted: 29 MAR 2005
- Manuscript Received: 16 MAR 2005
- health knowledge;
- health education;
- patient education;
- attitude to health
Ethnic and racial disparities in cancer outcomes are often attributed to sociodemographic factors, including income, education, insurance status, and access to quality cancer care. Less attention has been directed toward other possible contributors, such as misconceptions about cancer and its treatment.
Nine hundred fifty-seven U.S. adults who reported that they had never been diagnosed with cancer were identified and surveyed using a random digit dialing telephone survey. Sociodemographic associations with endorsement of five misconceptions about cancer treatment were evaluated by weighted logistic regression.
The most prevalent misconception, “Treating cancer with surgery can cause it to spread throughout the body,” was endorsed as true by 41% of the respondents. The second most prevalent misconception, “The medical industry is withholding a cure for cancer from the public in order to increase profits, ” was identified as true by 27%. The statement, “Pain medications are ineffective against cancer pain, ” was accurately rejected by 68% of the respondents, and 89% of the respondents correctly disagreed with the statement, “All you need to beat cancer is a positive attitude, not treatment. ” Eighty-seven percent of the respondents were able to identify the statement, “Cancer is something that cannot be effectively treated,” as false. Respondents who were older, nonwhite, Southern, or indicated being less informed about cancer endorsed the most misconceptions.
The prevalence of at least three of the five cancer misconceptions was unacceptably high, and varied by several sociodemographic factors. These beliefs may increase the risk for cancer morbidity and mortality because of poor adherence to treatment regimens. Cancer 2005. © 2005 American Cancer Society.
A recent review of cancer control progress in the United States reported an improvement in 5-year relative survival rates for all cancer types combined among males and females, respectively, from 42.4% and 56.6% during 1975–1979, to 63% and 64% twenty years later. The report also noted a continued decline in cancer mortality during that same time period.1 However, a growing body of research has documented persistent racial, ethnic, and socioeconomic disparities in cancer incidence rates, stage distributions, survival rates, and mortality rates.2, 3 Major contributing factors to cancer disparities include differences in access to and use of high-quality health care, including preventive and screening services as well as treatment. As noted in the 2003 Institute of Medicine report,4 social inequities can influence the interactions between patients and health care providers, and cultural factors can influence health behaviors, beliefs about health and illness, and approaches to medical care.
Several social theoretic frameworks, such as social cognitive theory, the health belief model, and the theory of planned behavior, have been proposed to explain and predict how patients make decisions about their health.5 Although these theories differ in some respects, all include the value of various health states and beliefs about the direction and magnitude of associations between health-related behaviors and health outcomes.
Sociodemographic variations in health literacy, attitudes, and beliefs can result in suboptimal health decisions about disease prevention, early detection, and treatment that contribute to health disparities.6–8 Members of certain racial and ethnic minority populations, individuals with less education, and the elderly are more likely than the general population to have a more fatalistic view of cancer and, consequently, to be less likely to take advantage of health care opportunities.9–13 For example, 1 study found that African Americans were approximately twice as likely as whites (61% vs. 29%) to believe that resection of lung carcinoma caused spread of disease and, for this reason, were more likely than whites to express concerns about undergoing surgery recommended by their physician.14
To address these issues, we undertook a study to assess the prevalence of beliefs regarding several common misconceptions about cancer treatment, and to identify the demographic factors associated with these dysfunctional beliefs. Such information is essential to gaining a better understanding of the basis for health disparities, and for designing, targeting, and implementing interventions in patient education.
MATERIALS AND METHODS
During early 2002, the American Cancer Society (ACS), Prevention magazine (published by Rodale Press, Emmaus, PA), and the Discovery health channel collaboratively designed and conducted a survey of cancer-related knowledge and beliefs called, “Cancer Myths and Cancer Facts.” The goal of this survey was to identify topics for collaborative public education projects/programming and for research. The survey collected demographic data and assessed attitudes regarding cancer risk factors, prevention, early detection, and treatment.
The questionnaire was developed by Princeton Survey Research Associates (PSRA) in collaboration with Rodale Press and the ACS research staff. To ensure the quality of the data, the questionnaire was pretested with a small group of respondents. The pretest interviews were monitored by PSRA staff and conducted using experienced interviewers who could best judge the quality of the answers given and the degree to which respondents understood the questions. Some final changes were made to screening procedures, question wording, and question order based on the monitored pretest interviews. The final questionnaire included questions regarding cancer risk factors, diagnosis, treatment, and prognosis, as well as information on age, gender, race, education, income, and other demographic factors. Data presented in the current article focus on the five items that assessed participants' misconceptions about cancer treatment and prognosis. The five misconception items were worded as statements, and the response categories were True, False, or Don't Know. The five misconception statements include the following: 1) “Pain medications are not effective in reducing the amount of pain people have from cancer”; 2) “All you need to beat cancer is a positive attitude, not treatment”; 3) “Treating cancer with surgery can cause it to spread throughout the body”; 4) “There is currently a cure for cancer but the medical industry won't tell the public about it because they make too much money treating cancer patients”; and 5) “Cancer is something that cannot be effectively treated.”
The sample was designed to represent the continental U.S. adult population in telephone households and was drawn using standard list-assisted random digit dialing methodology. Telephone interviews were conducted from November 4, 2002 to November 17, 2002 by trained interviewers. Up to a total of 10 attempts were made to contact individuals at every sampled telephone number. Calls were staggered over times of day and days of the week, with at least one daytime call. When household contact was established, interviewers asked to speak with the youngest adult (age ≥ 18 years) male currently at home. If no adult male was available, interviewers asked to speak with the oldest female at home. This systematic respondent selection technique has been shown to produce samples that mirror closely the population in terms of age and gender.15 Of the 3338 working telephone numbers (excluding fax, business, and nonworking phones) called, 2497 were reached successfully. Thus, the contact rate was 75%. Altogether, 1254 individuals agreed to participate, resulting in a cooperation rate of 50%. Of these 1254 respondents, 1070 were eligible (adults without language barriers that preclude participation) and 1002 actually completed the survey (94%). The overall response rate of 35% was calculated as the product of 3 components: the contact rate (75%), the cooperation rate (50%), and the survey completion rate (94%). Of the 1002 respondents, 45 reported having had a previous cancer diagnosis. As it might be expected that individuals with a history of cancer would have different knowledge and attitudes about cancer treatment compared with the general population, these individuals were removed from subsequent analyses. All data analyses henceforth were conducted with a sample of 957 U.S. residents with no history of cancer.
The sample of 957 adults was weighted to match national parameters for gender, age, education, race, Hispanic origin, and U.S. census region. These parameters were obtained from the March 2001 Current Population Survey, which included households with a telephone in the continental United States.15 The use of these weights in statistical analysis ensures that the demographic characteristics of the sample closely approximate the demographic characteristics of the national population. The unweighted and weighted distributions for gender, race/ethnicity, and age differed by no more than 2% in any category. For census region, the greatest difference was 2.5%. Weighting had a somewhat greater impact on education, with the unweighted distribution demonstrating 5.3% fewer individuals with less than a high school and 7.5% more college graduates. To compensate for loss in statistical efficiency resulting from departures from simple random sampling, we used a design effect (=1.10) to adjust tests of statistical significance.15
Weighted logistic regression was used to evaluate the association between belief in each misconception and several sociodemographic characteristics: gender, race/ethnicity (white, African American, Hispanic, other), age (< 35, 35–64, ≥ 65 years), education (less than high school education, high school graduate, some college, college graduate), annual household income (< $30,000, $30,000–49,999, $50,000–99,999, ≥ $100,000), family history of cancer (yes, no), self-reported general cancer knowledge (very informed, somewhat informed, not very informed, not at all informed), and U.S. census region (South, Northeast, Midwest, West). For statistical analyses, we included people with uncategorical responses (Don't Know or Refused) with the referent group (defined in Table 1), which was generally the largest group. In addition, because both True and Don't Know responses indicate low health literacy, we combined these responses for logistic regression analyses.
|Characteristics||Percentagea||Mean misconception score||Univariate P valueb||Multivariate P valueb|
|Less than high school||13.4||1.8||0.007||0.071|
|High school graduate||37.0||1.5||Referent||Referent|
|Family history of cancer|
|Not very informed||10.5||1.4||0.159||0.820|
|Not at all informed||1.6||2.1||0.000||0.013|
We evaluated the effect of each characteristic in univariate analyses and in multivariate analyses including all characteristics. All 95% confidence intervals were adjusted by the design effect. We also calculated a combined misconception score as an indicator of overall cancer treatment-related health literacy. We assigned each Not True response a zero value, each Don't Know response a value of 0.5, and each True response a value of 1. Thus, this score had a range of 0 (did not believe any of these misconceptions) to 5 (believed all five of these misconceptions). We used univariate and multivariate linear regressions to assess the association between the score and the sociodemographic characteristics. P values were adjusted by the design effect.
After weighting, our sample was representative of the 2000 U.S. population (US Census 2000) with regard to gender, age distribution, race/ethnicity, income, educational attainment, and census region (Table 1). Table 1 describes the sample characteristics weighted to the national standard population. The majority of participants (56%) reported a positive family history of cancer. A majority of participants said they were very or somewhat informed about cancer. Only 11 % said they were not very well informed, and < 2% said they were not at all informed.
Endorsement of Cancer Misconceptions
Only one-fourth of the respondents did not believe in any of the five cancer treatment misconceptions. The mean misconception score (1.3 overall) differed by most sociodemographic characteristics, with higher education and income levels being associated with more accurate responses. In contrast, people who were nonwhite, ≥ 65 years, residents of the South, and those who reported being not very or not at all informed about cancer had the highest misconception scores (Table 1).
The most prevalent misconception among the five statements was the third statement (Treating cancer with surgery can cause it to spread throughout the body). Almost 41% of the participants believed this statement to be true and an additional 13% indicated that they did not know. Only 46% correctly identified this statement as false (Table 2). Men were more likely than women to agree with this statement, as were people of other race/ethnicity and people who reported being not at all informed about cancer. People with at least some college education and those in higher income brackets were most likely to say that the statement that surgery could spread cancer was not true (Table 3).
|Misconception||True (%)||Don't know (%)||Not true (%)|
|1. Pain medications are not effective in reducing the amount of pain people have from cancer.||19.0||13.2||67.8|
|2. All you need to beat cancer is a positive attitude, not treatment.||6.6||4.5||89.0|
|3. Treating cancer with surgery can cause it to spread throughout the body.||40.8||12.9||46.3|
|4. There is currently a cure for cancer but the medical industry won't tell the public about it because they make too much money treating cancer patients.||27.3||14.3||58.5|
|5. Cancer is something that cannot be effectively treated.||7.0||5.7||87.3|
|Characteristics||Misconception 1 (ineffective pain medications)||Misconception 2 (positive attitude is all you need)||Misconception 3 (surgery spreads cancer)||Misconception 4 (cancer cure exists)||Misconception 5 (cancer cannot be treated)|
|Agree (%)b||ORc||ORd (95% CI)e||Agree (%)b||ORc||ORd (95% CI)e||Agree (%)b||ORc||ORd (95% CI)e||Agree (%)b||ORc||ORd (95% CI)e||Agree (%)b||ORc||ORd (95% CI)e|
|Women||31||1.00||1.00 (referent)||10||1.00||1.00 (referent)||51||1.00||1.00 (referent)||42||1.00||1.00 (referent)||13||1.00||1.00 (referent)|
|Men||34||1.14||1.20 (0.96–1.50)||12||1.32||1.38 (0.98–1.93)||56||1.24||1.29 (1.04–1.59)||41||0.98||1.02 (0.83–1.26)||13||1.02||0.95 (0.70–1.30)|
|White||28||1.00||1.00 (referent)||9||1.00||1.00 (referent)||51||1.00||1.00 (referent)||39||1.00||1.00 (referent)||11||1.00||1.00 (referent)|
|African American||37||1.43||1.15 (0.80–1.65)||13||1.47||1.26 (0.74–2.17)||60||1.41||1.23 (0.86–1.75)||49||1.45||1.28 (0.91–1.80)||11||0.95||0.80 (0.46–1.37)|
|Hispanic||40||1.63||1.59 (1.11–2.28)||19||2.39||1.93 (1.19–3.15)||55||1.12||1.07 (0.76–1.52)||39||0.99||0.91 (0.64–1.29)||21||2.04||1.81 (1.16–2.83)|
|Other||60||3.52||3.82 (2.28–6.41)||24||2.96||2.13 (1.10–4.13)||70||1.85||1.92 (1.12–3.29)||63||2.95||3.13 (1.86–5.29)||25||2.17||1.97 (1.05–3.72)|
|< 35||31||1.00||1.00 (referent)||12||1.00||1.00 (referent)||53||1.00||1.00 (referent)||41||1.00||1.00 (referent)||16||1.00||1.00 (referent)|
|35–64||30||0.93||1.15 (0.89–1.48)||8||0.56||0.69 (0.47–1.01)||51||0.88||0.93 (0.74–1.18)||39||0.89||0.94 (0.74–1.2)||11||0.65||0.73 (0.52–1.03)|
|≥65||43||1.64||1.63 (1.15–2.33)||18||1.42||1.31 (0.80–2.13)||65||1.59||1.29 (0.91–1.84)||50||1.42||1.21 (0.86–1.71)||12||0.74||0.75 (0.45–1.25)|
|Less than high school||44||1.38||1.19 (0.85–1.66)||19||1.75||1.54 (0.96–2.46)||69||1.32||1.22 (0.86–1.72)||58||1.55||1.5 (1.08–2.08)||19||1.58||1.44 (0.93–2.25)|
|High school graduate||36||1.00||1.00 (referent)||11||1.00||1.00 (referent)||63||1.00||1.00 (referent)||47||1.00||1.00 (referent)||13||1.00||1.00 (referent)|
|Some college||33||0.85||0.95 (0.72–1.27)||10||0.81||0.93 (0.59–1.46)||49||0.56||0.58 (0.44–0.76)||39||0.71||0.75 (0.57–0.98)||12||0.86||0.76 (0.51–1.15)|
|College graduate||19||0.42||0.53 (0.38–0.73)||8||0.62||0.99 (0.60–1.64)||36||0.33||0.35 (0.26–0.46)||27||0.41||0.45 (0.34–0.61)||10||0.73||0.76 (0.49–1.18)|
|< $30,000||42||1.44||1.32 (1.01–1.73)||16||1.29||1.27 (0.86–1.86)||60||1.06||0.86 (0.66–1.13)||49||1.39||1.18 (0.91–1.53)||14||1.21||0.99 (0.67–1.46)|
|$30,000–$49,999||29||1.00||1.00 (referent)||9||1.00||1.00 (referent)||58||1.00||1.00 (referent)||38||1.00||1.00 (referent)||10||1.00||1.00 (referent)|
|$50,000–$99,999||23||0.60||0.76 (0.56–1.03)||4||0.25||0.33 (0.18–0.60)||46||0.61||0.77 (0.59–1.01)||38||0.88||1.10 (0.84–1.45)||14||1.21||1.44 (0.96–2.16)|
|≥ $100,000||26||0.72||0.95 (0.61–1.47)||8||0.61||0.74 (0.37–1.45)||42||0.51||0.73 (0.50–1.09)||27||0.54||0.74 (0.48–1.12)||7||0.58||0.70 (0.35–1.42)|
|Family history of cancer|
|Yes||28||1.00||1.00 (referent)||7||1.00||1.00 (referent)||51||1.00||1.00 (referent)||39||1.00||1.00 (referent)||12||1.00||1.00 (referent)|
|No||37||1.53||1.34 (1.07–1.67)||16||2.59||2.21 (1.56–3.13)||56||1.25||1.12 (0.90–1.38)||44||1.24||1.14 (0.92–1.41)||13||1.10||0.97 (0.71–1.33)|
|Very informed||30||1.00||1.00 (referent)||11||1.00||1.00 (referent)||53||1.00||1.00 (referent)||42||1.00||1.00 (referent)||11||1.00||1.00 (referent)|
|Somewhat informed||32||1.09||1.04 (0.81–1.32)||8||0.78||0.69 (0.47–1.02)||53||0.99||0.96 (0.77–1.21)||41||0.95||0.93 (0.74–1.17)||12||1.10||1.06 (0.75–1.51)|
|Not very informed||36||1.33||1.16 (0.79–1.71)||18||1.82||1.34 (0.80–2.23)||56||1.13||0.90 (0.62–1.30)||42||0.99||0.85 (0.59–1.23)||18||1.81||1.57 (0.95–2.57)|
|Not at all informed||44||3.38||2.49 (1.27–4.91)||33||6.09||3.98 (1.96–8.07)||78||4.02||3.04 (1.32–6.99)||54||1.90||1.46 (0.75–2.84)||53||5.92||6.19 (3.06–12.5)|
|South||37||1.00||1.00 (referent)||10||1.00||1.00 (referent)||57||1.00||1.00 (referent)||45||1.00||1.00 (referent)||16||1.00||1.00 (referent)|
|Northeast||23||0.51||0.57 (0.41–0.80)||9||0.92||1.00 (0.60–1.66)||55||0.90||1.09 (0.81–1.47)||38||0.77||0.90 (0.67–1.21)||12||0.68||0.66 (0.43–1.01)|
|Midwest||32||0.80||0.96 (0.72–1.29)||10||0.96||1.18 (0.73–1.90)||51||0.76||0.83 (0.63–1.10)||41||0.86||0.99 (0.75–1.31)||9||0.53||0.54 (0.35–0.83)|
|West||31||0.76||0.76 (0.56–1.03)||15||1.62||1.66 (1.07–2.58)||50||0.74||0.86 (0.64–1.15)||39||0.78||0.91 (0.68–1.22)||12||0.68||0.65 (0.43–0.99)|
The next most commonly held misconception was the fourth statement (There is currently a cure for cancer but the medical industry won't tell the public because they make too much money treating cancer patients). More than one-fourth (27%) of the people in this survey agreed with this statement, 14% indicated that were not sure, and 59% correctly believed this statement to be false (Table 2). Except for education and other race/ethnicity, belief in this conspiracy theory was similarly common across all levels of each demographic characteristic. Compared with people who graduated from high school, people with less than a high school education were more likely to agree with this statement, whereas people who attended college were more likely to disagree (Table 3).
Nineteen percent of participants endorsed the first misconception statement (Pain medications are not effective in reducing the amount of pain people have from cancer) whereas 13% indicated that they did not know and 68% identified this statement as false (Table 2). The most marked difference was by the sociodemographic characteristic race/ethnicity. Compared with whites, Hispanics and individuals of other race/ethnicity were more likely to consider pain medications ineffective in treating cancer pain. Older and poorer individuals also were more likely to consider pain medications ineffective, as were people who reported no family history of cancer and those who said they were not at all informed about cancer. College graduates and residents of the northeastern United States were more likely to judge this statement as incorrect (Table 3).
Few people (7%) believed the second misconception statement (All you need to beat cancer is a positive attitude, not treatment) and 5% responded that they did not know. The majority (89%) of participants identified this statement as false (Table 2). This misconception was most prevalent among those who said they were not at all informed about cancer (33%). Hispanic and other racial/ethnic groups also were more likely to agree with this statement, as were people who reported no family history of cancer and residents of the western United States (Table 3).
Most people (87%) correctly disagreed with the fifth misconception statement (Cancer is something that cannot be effectively treated), but a modest percentage (7%) indicated that they believed this statement was true and an additional 6% reported that they did not know (Table 2). Hispanics and people with low self-rated cancer knowledge were the most likely to agree with this statement, whereas residents of the midwestern United States were least likely to agree (Table 3).
This national telephone survey of 957 U.S. adults indicated that many people harbor ≥ 1 misconceptions regarding cancer treatment. If these misconceptions are not modified through patient awareness and education programs, individuals may make inappropriate health care decisions.
Previous research has shown that approximately one-fourth of people recently diagnosed with cancer experience pain and that the prevalence of cancer pain increases to three-fourths of patients who have advanced cancer.16 Fortunately, optimally prescribed medications and other treatment modalities are successful in controlling pain in > 80% of patients.17, 18 As illustrated by the results of our survey in which 32% of respondents agreed with or were uncertain about the first statement (Pain medications are not effective in reducing the amount of pain people have from cancer), many people do not realize that it is unnecessary for patients to endure inadequately controlled cancer pain.
Other barriers may also prevent optimal pain control, especially for poorer individuals, members of medically underserved racial minorities, and older persons.19, 20 For example, pharmacies in some low-income neighborhoods do not stock opioid medications, which limits the options for pain control for residents of these communities.21 Other barriers include cultural differences in how pain is discussed with clinicians and racial/ethnic biases among some clinicians regarding drug-seeking behavior.22, 23 Consequently, the degree to which personal or community experience with suboptimal pain control perpetuates the misconception that pain medications are ineffective for cancer pain is important. Our results support the importance of patient education about pain control, so that all people living with cancer understand that pain is controllable and are aware of the medical options for relieving their pain.
Thirteen percent of respondents either agreed with the second misconception (All you need to beat cancer is a positive attitude, not treatment) or were uncertain whether this was true. In previous studies, 60% of survivors of breast carcinoma surveyed believed that a positive attitude could prevent a recurrence of their disease24 and 82.5% of survivors of ovarian carcinoma attributed their lack of cancer recurrence to a positive attitude, ranking it first among a host of other possible factors.25 Despite the finding that a large proportion of cancer survivors believe a positive attitude can help to improve their prognosis, results of empirical studies on this topic have been inconsistent. In one study, for example,26 a “fighting spirit” and denial were significantly associated with prolonged survival in women with breast carcinoma. In contrast, another study using the same questionnaire found no such correlation, but did find a modest relation between depression and helplessness/hopelessness and disease outcome.27 Overall, the evidence of a relation between a positive attitude and cancer survival is weak. A metaanalysis of 26 studies28 found little consistent evidence that psychological coping styles play an important role in survival from or recurrence of cancer. Regardless of its impact on cancer survival or recurrence, a positive attitude is likely still desirable for patients with cancer and survivors of cancer, as it might lead to improved treatment adherence as well as healthier lifestyle behaviors, such as increased physical activity, improved dietary patterns, and a reduction in tobacco use. However, patients and survivors should not be lulled into the belief that merely having a positive attitude will suffice to survive the illness or avoid disease recurrence. Conversely, they should not view cancer recurrence or progression as being due to their failure to maintain a positive attitude.
The most common misconception in our study was the third statement (Treating cancer with surgery can cause it to spread throughout the body). Only 46% of respondents recognized this statement as incorrect. Surgery is a highly effective treatment for localized cancer and an important part of multimodal therapy for many advanced-stage cancers.29 Approximately 75% of patients with cancer undergo a major surgical procedure as a part of their treatment, and nearly all have undergone a biopsy before definitive surgical, medical, or radiation therapy.29, 30 Certain surgical procedures have been shown to increase the risk of local spread of some cancers, such as ovarian and testicular cancers. However, surgical oncologists are aware of these issues and plan surgical therapy of these diseases to minimize this risk.29, 31, 32 A probable origin of this misconception is that until the past several decades, most cancers were detected at an advanced stage when prognosis was very unfavorable. Without modern imaging techniques, patients often underwent exploratory surgery, after which their poor outcomes were attributed to the surgery rather than to the advanced stage disease at the time of diagnosis. The consequence of this misconception is that patients who believe this may forego or delay surgery that might improve the length and/or quality of their lives.
Forty-two percent of individuals either believed or were unsure about the fourth statement (There is currently a cure for cancer but the medical industry won't tell the public about it because they make too much money treating cancer patients). There is little evidence to support this misconception. Furthermore, this conspiracy theory is not supported by progress in other diseases, such as tuberculosis or polio, where relatively inexpensive antibiotics or immunizations have largely replaced more expensive supportive care. This belief also assumes that the biomedical research community is willing to remain silent about a secret cure, even though prompt publication of research results is clearly in the personal self-interest of investigators. The clinical implication of this misconception is an attitude of mistrust among a substantial population of patients. These individuals might be less likely to adhere to treatment regimens and more likely to resort to less efficacious alternative therapies. Despite the absence of evidence supporting this alleged conspiracy, one can speculate as to why some individuals might believe this misconception. Progress in clinical oncology has been somewhat slower, in comparison to the breakthroughs during the past century in preventing and treating infectious diseases. The distinction between cancers and infections requires some understanding of their biologic bases, and of how biomedical research is conducted. This explanation for the conspiracy misconception is supported by its strong association with education. In addition, African Americans were less likely than whites and Hispanics to reject this misconception, possibly influenced by unfavorable views of the U.S. health care system based on personal experiences and awareness of past injustices such as the Tuskegee syphilis study.33, 34
Thirteen percent of respondents agreed with or were uncertain about the fifth statement (Cancer is something that cannot be effectively treated). Indeed, this finding is consistent with the literature on cancer “fatalism,” which is the belief that death is inevitable when cancer is present.11 Cancer fatalism has been shown to be a barrier to participation in cancer screening, detection, and treatment.35–37 In general, these studies also have found that the likelihood of believing such fatalistic statements is higher among racial/ethnic minorities, those with less education, those of lower income, and older individuals.12, 38
It is noteworthy that no single profile of individuals with low health literacy completely explains the sociodemographic variations in the prevalence of these misconceptions. For example, African Americans in the current study were mistrustful about cancer surgery, as has been reported previously,14 whereas other misconceptions were more prevalent among Hispanics or individuals in the other race/ethnic category, which included Asian Americans.
No particular region within the United States was consistently more knowledgeable regarding these cancer myths than another. However, a few differences in belief of these misconceptions were demonstrated. Residents of the northeastern United States were least likely to be pessimistic about control of cancer pain, southerners were least likely to believe that a positive attitude could control cancer, and midwesterners were least likely to express fatalism about cancer treatment.
We observed significant differences in prevalence of endorsement among individuals who rated their own cancer health literacy as very informed, somewhat informed, or not very informed for only one of the five misconceptions in our survey. However, individuals who rated their knowledge as not at all informed were generally quite accurate when rating their own health literacy. They were significantly more likely to endorse four of the five misconceptions in the survey. Because it is not known whether patients who recognize their limited cancer literacy will ask questions of their providers or seek information from sources such as the ACS or the National Cancer Institute, educational interventions are needed to reduce cancer misconceptions. The prevalence of misinformation among those rating their cancer literacy as very informed, somewhat informed, or not very informed suggests the need to provide educational information for all people, not just those diagnosed with cancer.
Evidence suggests that cancer knowledge and literacy may influence secondary prevention, care-seeking behavior, and participation in treatment decision-making among those diagnosed.39, 40 It also has been shown that knowledge about cancer, risk factors, and cancer treatment can serve to give patients with cancer a sense of control.41 Chalmers et al.42 found that people who were able to meet adequately their information, support, and communication needs appeared to have more reasonable perceptions of their risk of cancer, practiced more preventive self-care, and had a better adjustment to the risk of cancer. Conversely, information that exaggerates risk, is baseless, or does not have an actionable message may generate unnecessary anxiety.
Because of the abundance of information available from media, health professionals, and personal support systems, cancer awareness is increasing. However, the accuracy of the information about cancer varies widely from source to source, and even if the information given is correct, misunderstandings in the delivery may lead to faulty impressions. What is becoming more clear through research is that the information that we provide must clearly state the risks and benefits of treatment and be structured in a format that facilitates appropriate decisions.43, 44
Although the cross-sectional design and self-reporting of data are potential limitations in the current study, we can still conclude that there is substantial demographic variation in health literacy related to cancer treatment. These results indicate that public and patient education interventions are most urgently needed in cancer centers, medical practices, and other community organizations that serve large numbers of patients with these at-risk demographic characteristics.
The authors thank Dana Flanders, M.D., Ph.D., for advice regarding statistical analyses and Corinne Crammer, M.M., M.Div., for editorial advice and assistance. The authors acknowledge the intramural funding of the American Cancer Society, as well as outside funding from the Discovery Health Channel and Prevention Magazine.
- 3The unequal burden of cancer: an assessment of NIH research and programs for ethnic minorities and the medically underserved. Washington, DC: National Academy Press, 1999., .
- 4SmedleyBD, StithAY, NelsonAR, editors. Unequal treatment: confronting racial and ethnic disparities in health care. Washington, DC: National Academy Press, 2003.
- 5Factors affecting health behavior. In: YarbroCH, GoodmanM, FroggeMH, GroenwaldSL, editors. Cancer nursing principles and practices. 5th ed. Boston: Jones and Bartlett, 2000: 85–99., .
- 6Nielsen-BohlmanL, PanzerAM, KindigDA, editors. Health literacy. A prescription to end confusion. Washington, DC: National Academy Press, 2004.
- 16National Comprehensive Cancer Network. Cancer pain. National Comprehensive Cancer Care Network. Updated 2004. Available from URL: http://www.nccn.org/professionals/physician_gls/PDF/pain.pdf [accessed November 10, 2004].
- 25The views of ovarian cancer survivors on its cause, prevention, and recurrence. Med Gen Med. 2001; 3(4) [(formerly published in Medscape Women's Health, eJournal, 2001, 6(5)]. Available from URL: http://www.medscape.com/newarticle/408950 [accessed November 10, 2004]., , , , .
- 29Principles of surgical oncology. In: KufeDW, PollockRE, WeichselbaumRR, et al., editors. Cancer medicine. 6th ed. Hamilton, ON: B.C. Decker, Inc., 2003: 570–583., .
- 30Clinical oncology in a changing healthcare environment. In: KufeDW, PollockRE, WeichselbaumRR, ditors. Cancer medicine. 6th ed. Hamilton, ON: B.C. Decker, Inc., 2003: 1187–1191., .
- 31Ovarian cancer. In: KufeDW, PollockRE, WeichselbaumRR, et al., editors. Cancer medicine. 6th ed. Hamilton, ON: B.C. Decker, Inc., 2003: 1831–1862.,
- 32Testis cancer. In: KufeDW, PollockRE, WeichselbaumRR, et al., editors. Cancer medicine. 6th ed. Hamilton, ON: B.C. Decker, Inc., 2003: 1747–1768., , , et al.
- 35Fatalism reconceptualized: a concept to predict health screening behavior. J Gender Cult Health. 1998; 3: 85–100., .
- 43Is information always a good thing? Helping patients make “good” decisions. Med Care. 2002; 40(9 Suppl ): V39–V44..