Colorectal cancer (CRC) is one of the most commonly occurring cancers in both men and women. 1 In 2000, approximately 13,200 new cases were diagnosed and 58,860 persons died of the disease. 2 At least 10% of the CRC burden (13,200 new cases) is attributed to a primary genetic factor. 3,4 Hereditary nonpolyposis colon cancer (HNPCC) and familial adenomatous polyposis (FAP) are the two main hereditary syndromes that predispose individuals to certain cancers, including CRC. 5,6 Genetic testing is currently available for these familial CRC syndromes; however, the state of the science regarding the genetics of CRC is in its infancy and, although genes have been identified, there are many genetic mutations that have not currently been identified. Consequently, a negative test for a gene mutation does not guarantee that an individual does not have a genetic mutation. In addition, any family history of CRC is viewed as a primary risk factor for developing CRC, regardless of the objective genetic risk. 3,4 Because family history of CRC is viewed as a primary risk factor for developing CRC, this study focused on the health beliefs and health promotion behaviors of individuals who may be at risk for developing CRC by virtue that they have a first-degree relative (FDRs) (parents, siblings, and progeny) with CRC.
The purpose of this study was to:
- Describe the health beliefs of FDRs of individuals with CRC.
- Describe screening intentions and/or behaviors of these FDRs.
- Examine the associations among age, gender, level of education, and the Health Belief Model (HBM) variables and an FDR’s participation in health maintenance visits (HMVs) with healthcare providers.
The variables examined in this study were chosen based on the HBM and a thorough review of related empirical literature. The review identified gaps in research related to degree of CRC risk perceived by family members of individuals with CRC and the health maintenance behaviors, including CRC screening of these individuals.
Genetic and Familial Risk for Colorectal Cancer
Several factors are likely to influence expression of a hereditary form of CRC, including environmental influences, interaction of other genes (modifier genes), different types of mutations, and lifestyle. 5,6 Consequently, there is no way to definitively know whether several cases of CRC are the direct result of a gene mutation, familial risk, or both, which may place FDRs at risk for developing CRC. The Amsterdam Criteria were developed in 1991 as minimal guidelines for the identification of HNPCC. However, 10%–15% of individuals diagnosed with CRC do not meet the Amsterdam Criteria for an HNPCC diagnosis, although they have family members with this disease. The Bethesda Guidelines were developed in 1996 in response to this issue. As a more liberal set of guidelines, they have significantly increased the number of individuals identified with CRC who may have relatives at risk for developing cancer. 7 However, there are many patients who do not meet either of these criteria for identification of a genetic disease, yet they demonstrate what appears to be a familial risk.
FAP accounts for only 1% of all colorectal cancers 8; the mean age for developing polyps is 16 years, and CRC usually develops at a mean age of 36 years. The early age of onset and the distinct presentation with more than 100 polyps make the diagnosis of this genetic disease more straightforward than HNPCC. Many individuals with CRC that cannot be specifically attributed to an HNPCC or FAP mutation at this time may be eventually identified as having a specific syndrome associated with an oncogene or family of oncogenes. However, currently, testing for genetic mutations associated with CRC for most individuals and their FDRs is not indicated. 7 However, these individuals may be at risk for developing CRC.
The American Society of Clinical Oncology (ASCO) developed a set of recommendations regarding genetic testing for CRC. 9 These recommendations outline the importance of taking a careful family history to facilitate counseling regarding familial risk. They also use the Amsterdam Criteria and the Bethesda Guidelines to help clinicians formulate recommendations for genetic testing and CRC screening. Genetic testing is, however, not generally recommended by ASCO for individuals with CRC and their FDRs, despite evidence of familial risk, unless they meet the Amsterdam Criteria and/or the Bethesda Guideline parameters. In addition there are many reasons an individual and/or family members may choose not to undergo genetic testing. Healthcare providers, including oncology nurses, should understand these recommendations for genetic testing, understand that FDRs of individuals with CRC may not meet these criteria, and know that these FDRs could still be at increased risk for developing CRC and should be instructed regarding CRC screening guidelines, including fecal occult blood (FOBT) testing and colonoscopy.
Colorectal Cancer Screening
FDRs of individuals with CRC who were diagnosed when they were younger than 50 years are at increased risk for developing cancer, and CRC screening recommendations for these FDRs should reflect this increased risk. 7,10 If the disease is detected early, surgical cure can be achieved in a high percentage of patients. The lack of early symptoms makes identification and screening of asymptomatic at-risk FDRs important in reducing the considerable mortality and morbidity associated with CRC when it is diagnosed in later stages.
Although cancer screening tests that are effective in reducing mortality are limited to the breast and cervix at this time, evidence exists for the efficacy of a few screening tests for other cancers, including colorectal. However, screening rates for CRC are still relatively low, despite efforts to increase public and professional awareness of the disease. In addition, determinants of CRC screening behavior have not yet been identified in any studies examining screening and health maintenance behaviors among individuals at average risk. 11,12 There is however, growing momentum for promotion of screening tests for CRC, especially sigmoidoscopy and colonoscopy by healthcare providers during routine HMVs.
The American Cancer Society (ACS) recommends yearly FOBT and colonoscopy every 3–5 years beginning at age 40 for FDRs of CRC patients. 13 The guidelines of the American Gastroenterology Association (AGA) state that screening should be the same as that for average risk groups but begin at age 40 instead of age 50 for FDRs or individuals with a strong family history of CRC or polyps. It also recommends that members of families with hereditary CRC syndromes, such as HNPCC or FAP, have closer surveillance beginning at puberty for FAP and at 20 to 30 years of age for HNPCC, dependent upon the age of presentation for the proband. 14 FDRs of individuals with CRC are not routinely identified and assessed for CRC risk, despite increased risk and screening recommendations. 15
The HBM is useful for research and theory building, and it provided the organizing framework for this study. 16 The HBM helps to unify unrelated findings from previous research, to provide a clear presentation of ideas, and to assess relationships among variables. This model assumes that individuals perceive that they are at risk because of their current pattern of behavior (perceived susceptibility and severity). It also assumes that individuals perceive that a specific behavior change may be beneficial because it results in a valued outcome at acceptable cost. Finally, the HBM assumes that individuals perceive that they are competent (self-efficacy or confidence) to overcome perceived barriers to taking action. 16 Examination of and understanding HBM variables is helpful in studying an individual’s perception of genetic risk for cancer and in providing a foundation for appropriate and effective interventions that will benefit at-risk individuals and their families (Figure 1).
Little research has examined the health beliefs or cancer screening intentions and health maintenance behaviors of FDRs of individuals with CRC or the variables that may be associated with these factors. It was identified that FDRs of individuals with CRC are not routinely identified and assessed for CRC risk, despite screening recommendations and possible increased risk. The lack of communication about familial CRC risk among physicians, patients, and their FDRs was also identified as a factor that accounts for an FDR’s lack of participation in health maintenance behaviors, including CRC screening. 17–21 FDRs may be at risk for developing cancer and, as science and genetic technology advance, these individuals may be identified as candidates for presymptomatic genetic testing. However, in the meantime, CRC screening is important for FDRs of individuals with CRC regardless of definitive genetic risk to identify and ultimately develop interventions that would increase CRC screening among FDRs of individuals with CRC and decrease morbidity and mortality from CRC.
Sample and Procedure
A nonexperimental survey design was chosen because of the lack of data available regarding the screening intentions and behaviors of FDRs of individuals with CRC. A cross-sectional sample of patients was identified in 1999 through the Pennsylvania Cancer Registry. Complete registry data were available through 1995; therefore, patients (n = 1087) between the ages of 18 and 60 years who were diagnosed with CRC in 1995 were identified as the sample population. Approval to conduct the study was granted by the university institutional review board, and an application for access to registry data was approved by the Pennsylvania Department of Health. After cases were identified through the registry, the investigator identified the physician of record and requested permission to contact registry patients. This is the standard registry procedure necessary to access registry patients in Pennsylvania who, in turn, provide or deny access to their FDRs. After the physician of record granted permission, study packets were mailed to registry patients. The packets included letters requesting that registry patients provide one FDR with an enclosed additional study packet marked “relative.” The FDRs who agreed to participate in the study (n = 90) were the actual study subjects. Figure 2 provides a graphic representation of the study design, and Table 1 provides a description of the sampling process. Data were collected during a 9-month period. Reminders were not sent to patients due to cost of mailing these notices.
SOCIODEMOGRAPHIC AND HEALTH-RELATED DATA
Data obtained on the Personal and Family History Form (PFHF) (Figure 3), which was developed by the investigator, were used to describe the registry patients and FDRs and were helpful in identifying sociodemographic variables associated with HBM variables and an individual’s participation in health maintenance activities. The PFHF had 21 items, and the responses were coded and labeled using SPSS 9.0. Gender, level of education, and age were the 3 demographic variables included in the analysis. These variables were identified in the literature as items associated with predicting health maintenance behaviors. 1,5,6,11,12,17 In addition, there were questions on this form regarding an individual’s participation in health promotion activities, including HMVs. An HMV is required for an individual to initiate CRC screening, 17 and this item was used as the dependent variable in this study.
The original dependent variable was a question asking if the FDRs would participate in CRC screening if identified at risk for the disease. However, 96% of FDRs reported that they would participate in screening if they were at risk for CRC, making it impossible to examine associations between HBM variables, age, gender, and level of education and an individual’s intention to participate in CRC screening. Another concern was that this response might represent a social desirability response set and therefore threaten the validity of the study. Evidence supporting this concern was identified after examining FDR data that indicated many did not participate in most health maintenance activities such as breast self-examination (BSE), mammography, prostate-specific antigen tests, and flu shots, as well as HMVs. A proxy variable for intention to screen was required, and whether an FDR participated in HMVs was consequently used as the outcome variable for the analysis.
The United States Preventive Services Task Force (USPSTF) recommended using HMVs as an opportunity for screening. 22 Individuals who schedule an HMV are more likely to participate in cancer screening activities because these activities are usually initiated during these visits. 22,23 The USPSTF data provided support for using participation in HMVs as a substitute for intention to participate in CRC screening, the original dependent variable in this study.
HEALTH BELIEFS REGARDING COLORECTAL CANCER
The Champion Health Belief Model Scale (CHBMS) was used to collect information regarding the health beliefs of FDRs related to CRC (Figure 4). Champion 24 developed items for the HBM constructs of susceptibility, seriousness, benefits, barriers, health motivation, and confidence. These items focused on BSE and perception of risk for breast cancer, but Champion stated that the scales developed from her HBM research could be used with substitution of a word or phrase for different types of cancer. 24
When Champion 24 developed the scale, she determined internal reliability for each subscale of the CHBMS using Cronbach’s alpha, which ranged from 0.60 to 0.78. Results of the test–retest correlations for this scale during development, using Pearson r, ranged from .47 to .86 (p < .001).
The investigator adapted the CHBMS (with Champion’s permission) for this research by altering a few questions and by substituting “colon cancer” for “breast cancer” in the wording for the questions in each of the 6 subscales. To assess content validity of this scale, the adapted items were sent to 3 experts for judgment of relevancy in measuring HBM constructs as applied to CRC. Champion also reviewed the revised scale. The reviewers did not question or eliminate any items and determined that content validity was adequate.
Descriptive statistics were used to profile the entire patient registry sample (Table 2) and to compare patients and FDRs who participated in the study.In addition, descriptive statistics on selected items from the PFHF are reported in Table 3.
Each independent variable was examined in a univariate fashion to assess its association with the dependent variable. These variables were then entered into a logistic regression model. The logistic regression model was built entering age, gender, level of education, and the subscale scores from the 6 CHBMS variables, into the model to determine significant predictors of an FDR participating in HMVs. The level of significance was set at 0.05. Logistic regression analysis was used to test the combined influence of all variables on participation in HMVs (Table 4). The independent variables were entered for logistic regression analysis as a group, levels of significance were examined, and the best fit model was systematically determined by eliminating nonsignificant variables step by step. This method was used to determine the likelihood of each of the independent variables contributing to the prediction of an FDR’s participation in HMVs. The maximum likelihood method was used and the correlation matrix was examined to determine if there were any collinearity problems. At each step the −2 log likelihood and the P values were examined and a determination was made from these data how to proceed with the model building. A Chi-squared probability with one degree of freedom, and a .05 level of significance (3.841) was used for the −2 log likelihood determination. The logistic regression model was built by dropping variables from the model in relation to their level of significance starting with the highest value or least significant variable being dropped first. The investigator continued this method of analysis, dropping the least significant variable at each step of the logistic regression analysis until a final model was determined with 3 significant variables remaining out of the original 9 variables (Table 5). Odds ratios were also calculated, which indicate the probability of each significant independent variable in the final model predicting participation in HMVs.
The logistic regression analysis resulted in a model (Table 5) with perceived barriers, perceived seriousness, and level of education positively associated with an FDR’s participation in HMVs. Thus, FDRs of individuals with CRC were more likely to participate in HMVs if they did not perceive that there were too many barriers to CRC screening, if they perceived that CRC was a serious disease, and if they had a college or graduate degree.
To interpret the odds ratio for perceived barriers, other variables were considered at a fixed level. The odds ratio of 2.39 for perceived barriers indicated that the likelihood of an FDR not participating in HMVs was 2.39 times higher than the likelihood of participating for every unit increase on the 5-point CHBMS, if the other variables were held constant. That is, the more screening barriers perceived by the FDR, the less likely he or she was to participate in HMVs. The odds ratio of 0.41 for perceived seriousness indicated that an FDR was 41% less likely to not participate in HMVs for every unit increase on the 5-point CHBMS. That is, the more serious the FDR perceived CRC to be, the better the chance was that the FDR would participate in HMVs. Finally, the odds ratio of 0.31 indicated that the likelihood of an FDR not participating in HMVs was 31% less if the FDR had a college or graduate degree. That is, the more educated the FDR was, the more likely that individual was to participate in HMVs. The odds ratios and confidence intervals for the final model are outlined in Table 6.
This study provided valuable information regarding the participation of FDRs of patients with CRC in HMVs. In addition, this study furthered understanding of the health beliefs of FDRs of individuals with CRC and identified HBM variables and demographic characteristics that were associated with an FDR’s participation in HMVs and consequently cancer screening activities. This study also provided support for using the HBM to examine the participation of FDRs of individuals with CRC in HMVs.
The HBM proposes that the likelihood an individual will take a health-related action (intention) is determined by both the individual’s psychologic state of readiness to take action and by the perceived benefit of the action weighed against the perceived barriers involved in the proposed action. Readiness is determined by both perceived susceptibility to a particular disease and perceived severity of the consequences of developing the disease. 25 The action does not occur unless the individual believes in both personal susceptibility and the serious repercussions of the disease. 25 Sixty-seven percent of FDRs indicated that they participate in HMVs. It is possible that those FDRs who did not participate in HMVs (33%) did not believe they were at risk for CRC; the final logistic regression model supports this premise. Although most of the FDRs in this study believed that CRC was a serious disease (Tables 5 and 6) FDRs did not believe that they were at risk for CRC. These results provided one explanation for the majority of FDRs claiming that they would participate in CRC screening if they perceived that they were at risk for the disease. However, that they did not participate in HMVs where screening would be initiated indicates that they may not have perceived themselves as susceptible to or at risk for developing CRC.
There were a few limitations to this study that warrant comment. A power analysis was conducted and identified that an n of 198 FDRs were needed for the proposed analysis. The potential registry sample consisted of 1,081 patients, and research packets were mailed to 959 CRC patients. The researcher anticipated a response rate that would provide an n equal to at least the required 198 subjects; however, despite the large potential sample size of the 1995 registry (n = 1,081), the patient response (n = 248, 26%) and actual participation rates (n=174, 18%) of registry patients who provided access to the study sample or FDRs (n = 90), were quite low. It is interesting to note, however, that the participation rate for FDRs who were contacted by their relatives with CRC (n = 174) and asked to participate in the study, was 50%. The response and participation rates in this study may have reflected privacy concerns of registry patients and the reluctance of some individuals to discuss cancer and genetic risk. Because inherited diseases are family diseases, 26 individuals with CRC may be reluctant to involve relatives in a study that could potentially identify these individuals at risk for cancer. Also, issues surrounding the privacy of an individual and the responsibility of that individual to warn family members of potential risks can create family tension. In addition, the possibility of a familial or genetic component in cancers experienced by the family may place that family and its members at risk for insurance and/or job discrimination. These issues may also create barriers to participation in research studies related to genetic risk.
Another limitation was that minority at-risk individuals were under-represented. There were 113 minority patients (10%) in the original registry data. Of the 174 patients who agreed to participate in the study, 6 were black, 1 Native American, and 1 chose other (total 4%) and only 4 minority relatives (4% of the 90 FDRs) agreed to participate (3 black and 1 Native American). Because differences in survival between African Americans and whites can be attributed to later stage at diagnosis among African Americans, 27–29 future research must include initiatives to increase minority recruitment in studies that examine participation in HMVs that may lead to participating in cancer screening activities.
The use of a mailed survey may also be responsible for the low participation rates. Telephone or in-person interviews would be a better strategy for recruitment because since it would provide the potential study participant with the opportunity to ask questions. Finally, these study results cannot be generalized because the low participation rate may have prevented certain variables from being significant as a result of insufficient statistical power.
CRC is an important public health concern in the United States, and clinical research over the last decade has focused on a variety of areas ranging from refinements in diagnostic techniques to development of new chemotherapy drugs. The identification of patients at risk for CRC based on family history has improved monitoring and screening strategies. For example, promotion of the use of flexible sigmoidoscopy and colonoscopy in at-risk individuals is likely to result in CRC diagnosis at earlier stages and consequently decrease the morbidity and mortality associated with CRC diagnosed at later stages. 30
Despite that CRC is highly curable when detected early, a recent survey showed that almost half of all Americans older than 50 years have never been screened for the disease. 8 The goals set forth in Healthy People 2000 included that 50% of the population older than 49 years of age have had FOBT within the past 2 years and for 40% to have a sigmoidoscopy. 8 The perception that screening involves uncomfortable procedures, coupled with the embarrassing nature of colorectal diseases, may explain some of the reluctance people have surrounding CRC screening. In addition, according to a recent congressional report, women are among the most likely to avoid screening for CRC because they perceive it to be primarily a male disease. 8 It is of note that in this study gender was not a significant predictor of intention to screen.
Attempts to promote CRC screening have used a public health model, targeted entire communities (mass media campaigns), and a medical model, targeting individuals (general practice patients). 12 These efforts did not include a systematic evaluation of perceptions regarding CRC risk, seriousness, the benefits and barriers to screening, an evaluation of an individual’s perception of ability to participate in screening activities or the identification and screening of FDRs of individuals with CRC. In addition, these approaches did not include an evaluation of strategies that promote screening and increase adherence to screening. And finally, few programs base their strategies on behavioral science models or theories. 12
Increased consumer awareness about the benefits of screening for breast and prostate cancer led patients to request screening more frequently for these diseases. Perhaps a similar strategy would improve participation in CRC screening programs. The National CRC Roundtable (NCCR) is a group assembled by the ACS and the Centers for Disease Control and Prevention. This group is dedicated to reducing incidence of and mortality from CRC in the United States, and it has developed a multimedia campaign to increase public awareness about the benefits of CRC screening. Recommended strategies include a plan for physicians to incorporate ACS recommendations for early detection of CRC in asymptomatic people as a part of HMVs, the production and airing of public service announcements on television and radio, and development of an internet chat room on CRC that allows patients, family members, survivors, and the interested public to discuss issues related to CRC. 8 These strategies have the potential to be useful if they are culturally sensitive and based on the results of clinical research that identifies variables associated with cancer screening intentions and behaviors of at-risk individuals.
The importance of healthcare provider recommendations has been documented in a number of studies that examined mammography and CRC screening. Engaging patients in a discussion about CRC risk is important in patient compliance with and participation in CRC screening activities. 11 It is also important that oncology nurses take the initiative to recommend screening when appropriate for patients to actually participate in screening activities for CRC. 11 Physician and nurse recommendations regarding CRC screening for FDRs of individuals with CRC are not well documented. Given the potential influence of physician and nurse recommendations on subsequent behaviors, this area must be explored to help develop interventions aimed at increasing CRC screening among FDRs and decreasing morbidity and mortality from this disease.
Few studies have explored the health beliefs, risk perceptions, and associated variables that motivate or provide barriers to CRC screening for individuals at increased risk for developing CRC. The results of this study provide valuable information regarding the health beliefs and sociodemographic variables associated with an individual’s participation in HMVs.
Despite the limitations of this study, important information emerged when the researcher examined the situations that created the limitations. Many of the difficulties stemmed from the methods required when conducting research using a registry. In addition, genetics related information can be perceived as sensitive, resulting in individuals choosing not to participate in research studies related to genetic risk, especially when they receive the survey in the mail from a stranger. Thirty-eight registry patients called the investigator to ask how their data had been accessed. In general, they were unaware that data relating to their cancer diagnosis was in a state registry.
Examination of perception of risk, perceived seriousness of CRC, perceived barriers to screening, and the importance of education level in predicting participation in HMVs provided insight into why people behave in certain ways regarding health-related issues and decision-making processes. The results suggest that future studies should examine factors related to initiation of and compliance with CRC screening, including healthcare providers’ (nurses’ and physicians’) recommendations for CRC screening for FDRs of individuals with CRC, and components of interventions that will help to promote CRC screening among FDRs.
There are many issues that must be considered for individuals at risk for any cancer 31 because these issues will influence participation in research studies and consequently the outcome of these studies. At this time the influence of hereditary genetic factors on CRC development is not completely understood and a family history of CRC is recognized as a risk factor. Oncology nurses should have a solid knowledge base in genetics to evaluate family risk and make recommendations regarding screening activities for FDRs. Nurse researchers must also consider the broader social issues related to having what may be a genetic or “family disease”. Insurance and work discrimination, 31 confidentiality, and the ownership of genetic materials and information, are a few issues that must be recognized when planning research to better serve individuals diagnosed with cancer and their families. Oncology nurses are in key positions to spearhead the efforts to educate families regarding familial risk for cancer and cancer screening recommendations for their families.
The author thanks Dr Jennifer Pinto Martin and Dr Deborah B. McGuire for their comments on the manuscript.
1. Kinney AY, Choi YA, DeVellis B, Kobetz, E, Millikan RC, Sandler RS. Interest in genetic testing among first-degree relatives of colorectal cancer patients. Am J Prevent Med. 2000; 18:249–252.
2. Greenlee RT, Murray T, Bolden S, Wingo PA. Cancer statistics, 2000. CA Cancer J Clin. 2000; 50:7–33.
3. Lynch H, de la Chapelle A. Genetic susceptibility to non-polypos colorectal cancer [review]. J Med Genet. 1999; 36:801–818.
4. Lynch H, Lynch J. Hereditary non-polyposis colorectal cancer. Sem Surg Oncol. 2000; 18:305–313.
5. Burke W, Peterson G, Lynch P, et al. Recommendations for follow-up care of individuals with an inherited predisposition to cancer. JAMA. 1997; 227:915–920.
6. Lynch H, Lynch J. Genetic counseling for hereditary cancer. Oncology. 1996; 10:27–32.
7. Postier RG, Brandt EN. Colorectal cancer: genetics and screening. J Oklahoma State Med Assoc. 1999; 92:261–266.
8. American Cancer Society. CA news and views: colorectal cancer screening. CA Cancer J Clin.
9. American Society of Clinical Oncology. Public Issues Committee of the American Society of Clinical Oncology: genetic testing for cancer susceptibility. J Clin Oncol.
10. Giardiello FM. Genetic testing in hereditary colorectal cancer. JAMA. 1997; 278:1278–1281.
11. Beeker C, Kraft JM, Southwell BG, Jorgensen CM. Colorectal cancer screening in older men and women: qualitative research findings and implications for intervention. J Comm Health. 2000; 25:263–278.
12. Vernon S. Participation in colorectal cancer screening: a review. J Natl Cancer Inst. 1997; 89:1406–22.
13. Jessup JM, Menck HR, Fremgen A, Winchester DP. Diagnosing colorectal carcinoma: clinical and molecular approaches. CA Cancer J Clin. 1997; 47:70–92.
14. Briskey EN, Pamies RJ. Colorectal cancer: update on recent advances and their impact on screening protocols. J Nat Med Assoc. 2000; 92:222–230.
15. Lerman C, Hughes C, Trock B, et al. Genetic testing in families with Hereditary Nonpolyposis Colon Cancer. JAMA. 1999; 17:1618–1622.
16. Strecher V, Champion V, Rosenstock I. The health belief model and health behavior. In: Gochman D, ed. Handbook of Health Behavior Research I, Personal and Social Determinants. New York: Plenum Press; 1997:71–91.
17. Shvartzman P, Rivkind E, Neville A, Friger M, Sperber AD. Screening intention and practice among first-degree relatives of colorectal cancer patients in southern Israel. Isr Med Assoc J. 2000; 2:675–678.
18. Pho K, Geller A, Schroy PC. Lack of communication about familial colorectal cancer risk associated with colorectal adenomas. Cancer Causes Control. 2000; 11:543–546.
19. Beaudin DJ. Results of screening first-degree relatives of colorectal cancer: a community practice perspective. Can J Gastroenterol. 2000; 14:489–492.
20. Charalambopoulos A, Syrigos KN, Ho JL, Murday VA, Leicester RJ. Colonoscopy in symptomatic patients with positive family history of colorectal cancer. Anticancer Res. 2000; 20:1991–1994.
21. Rawl SM, Menon U, Champion VL, Foster JL, Skinner CS. Colorectal cancer screening beliefs. Focus groups with first-degree relatives. Cancer Pract. 2000; 8:32–37.
22. Cohen SJ, McClatchey MW, Wolfe P, Shelton BJ, Archer PG. Health maintenance visits as the determinant of cancer screening by primary care physicians. J Gen Intern Med. 1996; 11S(suppl):121.
23. Yarnell SH, Michener JL, Broadhead WE, Tse CKJ. Increasing compliance with mammography recommendations: health assessment forms. J Fam Pract. 1993; 36:59–64.
24. Champion V. Instrument development for health belief model constructs. Adv Nurs Sci. 1984; 6:73–85.
25. Mikhail B. The health belief model: a review and critical evaluation of the model, research, and practice. Adv Nurs Sci. 1981; 161:65–82.
26. Croyle R, Achilles J, Lerman C. Psychologic aspects of cancer genetic testing. Cancer Suppl. 1997; 80:569–575.
27. Report of the Secretary’s Task Force on Black & Minority Health. Vol I. Executive Summary.
28. Pennsylvania Department of Health. Pennsylvania Cancer Incidence and Mortality 1991–1995.
Harrisburg, Pa: Division of Health Statistics Pennsylvania Cancer Registry Pennsylvania Cancer Control Program; 1998.
29. Stromborg MF, Johnson J, McCorkle R. A program model for nurses involved with cancer education of Black Americans. J Cancer Educ. 1987; 2:145–151.
30. Cohen A. Colorectal cancer: evolving concepts in diagnosis, treatment, and prevention. CA Cancer J Clin. 1999; 49:199–201.
31. Giarelli E, Jacobs LA. Issues related to the use of genetic material and information. Oncol Nurs Forum. 2000; 27( 3):459–467.