Disparities in Cancer Incidence and Mortality Among Delaware Residents, 1998–2002 June 2006 DELAWARE HEALTH AND SOCIAL SERVICES DIVISION OF PUBLIC HEALTH Health Promotion and Disease Prevention Section This report is made possible with funding from the Delaware Health Fund and strategic leadership and guidance provided by the Delaware Cancer Consortium. CONTENTS EXECUTIVE SUMMARY 1. INTRODUCTION 1 1.1. Delaware Cancer Statistics 5 1.2. Disparities in Cancer Burden 5 1.2.1. Factors That Contribute to Disparities 6 1.3. Objectives 8 2. METHODS 9 2.1. Incidence and Mortality Data 9 2.1.1. Data Sources 9 2.1.1.1. Delaware Cancer Registry 9 2.1.1.2. Surveillance, Epidemiology, and End Results Program 9 2.1.1.3. National Center for Health Statistics 10 2.1.1.4. Census Data 10 2.1.2. Variable Definitions 11 2.1.3. Analytic Methods 11 2.1.3.1. Direct Standardization 11 2.1.4. Disparities in Incidence and Mortality 12 2.1.5. Trends in Cancer Incidence and Mortality 13 2.1.6. Data Reporting Rules 13 2.2. Behavioral Cancer Risk Factors and Screening Usage 13 2.2.1. Data Sources 13 2.2.2. Variable Definitions 13 2.2.3. Analytic Methods 14 2.3. Cancer Stage at Diagnosis 15 2.3.1. Data Sources 15 2.3.2. Variable Definitions 15 2.3.3. Analytic Methods 15 2.4. Cancer Treatment 16 2.5. Data Interpretation 17 3. RESULTS 19 3.1. Cancer Incidence and Mortality 19 3.1.1. Cancer Burden in Delaware 19 3.1.2. Disparities in Cancer Incidence and Mortality 19 3.1.3. Cancer Disparities by Sex, Age, and County of Residence 24 3.1.3.1. Disparities by Sex 24 3.1.3.2. Disparities by Age 27 3.1.3.3. Disparities by County of Residence 30 3.2. Trends in Cancer Incidence and Mortality 33 3.2.1. Cancer Incidence in Delaware, 1980–2002 34 3.2.1.1. All Cancer Sites Combined 34 3.2.1.2. Female Breast Cancer 35 3.2.1.3. Colorectal Cancer 36 3.2.1.4. Lung and Bronchus Cancer 37 3.2.1.5. Prostate Cancer 38 3.2.2. Cancer Mortality in Delaware, 1980–2002 39 3.2.2.1. All Cancer Sites Combined 39 3.2.2.2. Female Breast Cancer 40 3.2.2.3. Colorectal Cancer 41 3.2.2.4. Lung and Bronchus Cancer 42 3.2.2.5. Prostate Cancer 43 3.3. Factors That Contribute to Disparities 43 3.3.1. Cancer Risk Factors and Screening Usage 44 3.3.2. Stage at Diagnosis 58 3.3.2.1. Female Breast Cancer 58 3.3.2.2. Colorectal Cancer 61 3.3.2.3. Prostate Cancer 64 3.3.3. Cancer Treatment 65 3.4. Health Policy, Health System, and Societal Factors 68 REFERENCES APPENDIXES A: Supplemental Information on Report Methodology B: Supplemental Incidence and Mortality Data C: Supplemental BRFSS Information LIST OF TABLES Table 1. Comparison of Delaware With the SEER Population 10 Table 2. Number of New Cancers and Deaths From Cancer in Delaware, by Race/Ethnicity 19 Table 3. Incidence and Mortality Rate Ratios (With 95-Percent Confidence Intervals) Comparing African Americans With Whites in Delaware and the United States 23 Table 4. Incidence and Mortality Rate Ratios (With 95-Percent Confidence Intervals) Comparing African Americans With Whites, by Sex, in Delaware and the United States 25 Table 5. Incidence and Mortality Rate Ratios (With 95-Percent Confidence Intervals) Comparing African Americans With Whites, by Age, in Delaware 28 Table 6. Incidence and Mortality Rate Ratios (With 95-Percent Confidence Intervals) Comparing African Americans With Whites, by County of Residence, in Delaware 31 Table 7. Comparison of Demographic Characteristics Between Delaware and the United States, 2002 44 Table 8. Access to Health Care in Delaware and the United States, by Race/Ethnicity 45 Table 9. Comparison of Modifiable Risk Factors for Cancer Between Delaware and the United States, by Race/Ethnicity 47 Table 10. Comparison of the Percentage of Individuals Who Have Not Received Cancer- Screening Tests Within the Recommended Time Interval Between Delaware and the United States, by Race 49 Table 11. Association Between Race/Ethnicity and Access to Health Care in Delaware 51 Table 12. Multivariate Regression Model of Predictors of Lack of Access to Health Care 51 Table 13. Association Between Race/Ethnicity and Modifiable Risk Factors for Cancer in Delaware 53 Table 14. Multivariate Regression Model of Predictors of Having Modifiable Risk Factors for Cancer in Delaware 53 Table 15. Association Between Race and Not Receiving Cancer Screening Tests in Delaware 56 Table 16. Multivariate Model of Predictors for Not Receiving Cancer Screening Tests in Delaware 56 Table 17. Number and Percentage of Delaware Breast or Colorectal Cancer Patients With Local Disease Treated for Cancer in Delaware, by Receipt of Standard / Approrpriate Cancer Treatment 67 Table 18. Association Between Receipt of Standard Treatment and Treating Facility, Cancer Type and Race Among Individuals with Local Stage Disease and Treated in Delaware 67 LIST OF FIGURES Figure 1. A model for the development of health disparities 7 Figure 2. Age-adjusted incidence rates for all cancer sites combined, Delaware and the United States 20 Figure 3. Age-adjusted mortality rates for all cancer sites combined, Delaware and the United States 21 Figure 4. Age-adjusted incidence rates for the four major cancer sites, by race, Delaware and the United States 22 Figure 5. Age-adjusted mortality rates for the four major cancer sites, by race, Delaware and the United States 23 Figure 6. Age-adjusted incidence rate differences comparing African Americans with Whites for colorectal and lung and bronchus cancer, by sex, Delaware and the United States 26 Figure 7. Age-adjusted mortality rate differences comparing African Americans with Whites for colorectal and lung and bronchus cancer, by sex, Delaware and the United States 27 Figure 8. Incidence rate differences comparing African Americans with Whites for the four major cancers, by age, in Delaware 29 Figure 9. Mortality rate differences comparing African Americans with Whites for the four major cancers, by age, in Delaware 30 Figure 10. Incidence rate differences comparing African Americans with Whites for the four major cancers, by county of residence, in Delaware 32 Figure 11. Mortality rate differences comparing African Americans with Whites for the four major cancers, by county of residence, in Delaware 33 Figure 12. Trends in cancer incidence for all sites combined comparing Whites with African Americans, by sex, in Delaware 34 Figure 13. Trends in female breast cancer incidence comparing Whites with African Americans, in Delaware 35 Figure 14. Trends in colorectal cancer incidence comparing Whites with African Americans, by sex, in Delaware 36 Figure 15. Trends in lung and bronchus cancer incidence comparing Whites with African Americans, by sex, in Delaware 37 Figure 16. Trends in prostate cancer incidence comparing Whites with African Americans, in Delaware 38 Figure 17. Trends in cancer mortality for all sites combined comparing Whites with African Americans, by sex, in Delaware 39 Figure 18. Trends in female breast cancer mortality comparing Whites with African Americans, in Delaware 40 Figure 19. Trends in colorectal cancer mortality comparing Whites with African Americans, by sex, in Delaware 41 Figure 20. Trends in lung and bronchus cancer mortality comparing Whites with African Americans, by sex, in Delaware 42 Figure 21. Trends in prostate cancer mortality comparing Whites with African Americans, in Delaware 43 Figure 22. Racial/ethnic disparities in access to health care in Delaware and the United States 46 Figure 23. Racial/ethnic disparities in behavioral cancer risk factors in Delaware and the United States 48 Figure 24. Racial disparities in cancer screening test usage in Delaware and the United States 50 Figure 25. Percentage of female breast cancer cases, by stage at diagnosis and race, in Delaware 58 Figure 26. Percentage of female breast cancer cases diagnosed at an advanced stage, by race and county of residence, in Delaware 59 Figure 27. Percentage of female breast cancer cases diagnosed at an advanced stage, by race and age at diagnosis, in Delaware 60 Figure 28. Percentage of colorectal cancer cases, by stage at diagnosis and race, in Delaware 61 Figure 29. Percentage of colorectal cancer cases diagnosed at an advanced stage, by race and county of residence, in Delaware 62 Figure 30. Percentage of colorectal cancer cases diagnosed at an advanced stage, by race and age at diagnosis, in Delaware 63 Figure 31. Percentage of prostate cancer cases, by stage at diagnosis and race, in Delaware 64 Figure 32. Percentage of prostate cancer cases diagnosed at an advanced stage, by race and county of residence, in Delaware 65 Figure 33. Percentage of prostate cancer cases diagnosed at an advanced stage, by race and age at diagnosis, in Delaware 66 EXECUTIVE SUMMARY OF FINDINGS CANCER INCIDENCE AND MORTALITY The results of this investigation demonstrate that among all racial and ethnic groups in Delaware, African Americans have disproportionately high overall cancer incidence and mortality rates. Compared with Whites, African Americans have higher incidence rates for colorectal, lung and bronchus, and prostate cancer. Cancer mortality rates among African Americans are higher than those for Whites for each of the four cancer sites examined, with the greatest disparity in prostate cancer mortality. The patterns of these disparities in incidence and mortality for cancers of the female breast, colon, lung and bronchus, prostate, and all sites combined are consistent with the literature based on national data. Data trends in cancer incidence and mortality from 1980 to 2002, based on data provided by the Delaware Cancer Registry, show that cancer incidence is decreasing among Whites for all cancer sites examined, except that in recent years an increase in prostate cancer was observed. Among African Americans, incidence rates are decreasing for all cancers combined, female breast, and lung and bronchus cancer, but in recent years the incidence is increasing for colorectal and prostate cancer. The data show cancer mortality decreasing among all races for all cancer sites except female breast and colorectal cancer. Mortality rates among African Americans for all cancers combined, lung and bronchus cancer, and prostate cancer are decreasing faster than those among Whites, narrowing the disparity over time. However, the disparity in mortality rates between African American and White men for colorectal cancer is increasing over time. Factors That Contribute to Disparities The observation that the disparities in incidence, mortality, and time trends are not uniform across cancer sites suggests that there might be disparities in modifiable behavioral risk factors, screening, stage of disease at time of diagnosis, and treatment, which may differ by cancer site. Using BRFS data, we investigated whether there were any notable racial/ethnic differences in access to health care, the prevalence of modifiable cancer risk factors, and the use of screening. The data showed that a smaller percentage of African Americans and Hispanics have health insurance, a personal doctor, and a reliable source of medical care than Whites. It is encouraging to note that for all three measures, minorities in Delaware have greater access to health care than minorities in the rest of the United States. The prevalence of modifiable behavioral cancer risk factors differed by race/ethnicity. Fewer African Americans than Whites participated in leisure or occupational physical activity and ate five or more fruits and vegetables per day. African Americans had a greater prevalence of obesity than other races. Whites were the most likely to have ever smoked and chronically consumed alcohol. Hispanics had higher rates of obesity than Whites. Education had the strongest and most consistent impact on behavioral risk factors, with individuals with less than a high school education more likely to be at risk. Screening behavior was most influenced by having a personal doctor; Whites and African Americans were screened at essentially the same rates. There were no strong associations observed between race and stage of diagnosis. Initial analyses, limited to only Delaware Cancer Registry data for four cancers (female breast, colorectal, lung/bronchus and prostate) diagnosed at local stage of disease, suggested that fewer African Americans than Whites received the study-defined “standard” treatment for these cancers. A statistically significant difference was observed for colorectal cancer. Subsequent, refined analyses, expanded to include facility-based source data and to accommodate legitimate reasons for lack of “standard” treatment (e.g., a patient’s decision to decline further treatment, or the presence of co-morbid conditions precluding radiation therapy), demonstrated a statistically significant difference only for treating facility; i.e., there was a strong association between the treating facility and the likelihood of receiving “appropriate” treatment. This association could be due to differences in documentation, rather than differences in practice. Health Policy and Barriers to Care Research reveals the impact of patient and system barriers on racial/ethnic cancer health disparities in the United States. Patient barriers, such as poverty, lack of health insurance, and lack of health literacy, affect the delivery of cancer prevention and treatment services received by minorities and the medically underserved. System barriers prevent cancer services from being provided as a continuum of care and create an inequality in the delivery of cancer care. As these barriers continue to contribute to cancer health disparities in the United States, health care delivery systems and Federal, State, and local agencies need to take the initiative to address these barriers and create equality in cancer care for racial/ethnic populations. COMPARISON OF DISPARITIES IN DELAWARE VERSUS THE UNITED STATES Disparities in incidence in Delaware were comparable or smaller than were those in the United States. Mortality disparities were smaller in Delaware than in the United States with the exception of colorectal cancer; mortality rates were higher in Delaware for all cancer sites combined and lung and bronchus cancer. The general pattern for incidence observed in Delaware was seen in the United States; however, the incidence of prostate cancer among African American men was increasing in Delaware, while rates had stabilized nationally. Also, although an increase was observed in the United States in lung and bronchus cancer among African American women, the rate among African American women in Delaware has been declining. Patterns in mortality rates were similar between Delaware and the United States. The disparity in colorectal cancer mortality was widening in both Delaware and the United States. For female breast cancer, there was a similar pattern in Delaware and the United States, although Delaware had higher rates with less disparity. The general pattern for prostate cancer was similar over time; less disparity was observed in Delaware, and mortality rates were lower (3–5, 9). Based on data from the BRFS, the Delaware population had a higher percentage of African Americans but a lower percentage of Hispanics than the national average. Educational attainment in Delaware was comparable with national levels, while income was higher in Delaware than in the United States for all races. On all measures of access to health care examined, Delaware residents of all races were more likely to have access than the rest of the United States. There is, however, greater disparity in the prevalence of behavioral risk factors between African Americans and Whites in Delaware than in the United States. Though in Delaware we did not find significant differences in stage at diagnosis, disparities in this area have been documented in some studies (8). Disparities in treatment have been documented in other investigations for early stage female breast, colorectal, and lung and bronchus cancer (32, 43–45). CAUSES OF DISPARITY The NIH Strategic Plan to Reduce and Ultimately Eliminate Health Disparities has three major goals: 1) to increase minority health and disparities research, 2) to increase opportunities for disparity research training and career development, and 3) to increase outreach to ensure that the public, health care professionals, and communities are educated about the latest advances in disparities research. The development of health disparities is a complex process that involves biological, cultural, socioeconomic, and political factors. These in turn can affect health practices, psychosocial and environmental stress, psychosocial resources, and the medical care that is received. Reducing and eliminating these disparities therefore requires a multidisciplinary approach. Using available data, we chose to examine the components of cancer prevention as potential factors that contribute to disparity. Data from DCR and the BRFS were examined to determine whether hypotheses could be developed regarding disparities observed in primary, secondary, and tertiary prevention efforts with disparities observed in cancer incidence and mortality. The small disparity in colorectal cancer incidence could be related to behavioral risk factors, including exercise, diet, and obesity. The disparity in colorectal cancer mortality may be due to lower screening rates. Increased lung and bronchus cancer incidence rates observed among African Americans run counter to their lower smoking rates. The substantial disparity observed among African Americans with prostate cancer cannot be explained by either of these factors. African American men were slightly more likely to have cancer risk factors but were also more likely to be screened. They were less likely to have advanced disease; however, their prostate cancer incidence and mortality rates were substantially higher than those of other races. While female breast cancer screening rates between Whites and African Americans were comparable, African Americans were more likely to be diagnosed with advanced disease, which may be a contributing cause of higher breast cancer mortality among African American women. LIMITATIONS This investigation had two major limitations. The first was that there were insufficient numbers of cases to examine cancer incidence and mortality rates and time trends consistently for racial/ethnic groups other that Whites and African Americans. This was a function of the size of Delaware and its smaller Hispanic population relative to the national population (five versus 14 percent). The second limitation was the limited amount of data available from DCR. To make some inferences about the causes of disparity, we would need to examine behavioral cancer risk factors, screening usage, socioeconomic factors, health literacy, and patient and system barriers to health care in the population that developed cancer. This limitation is currently being address by a retrospective study of Delaware residents who were diagnosed with colorectal, female breast, lung and bronchus, and prostate cancer between 1999 and 2003. Study participants are interviewed regarding their demographics, access to health care, risk factors for cancer, screening history, and the cancer treatment received. 1. INTRODUCTION Cancer is a major public health burden in the United States. In Delaware, as in other States, this burden is not distributed equally along demographic lines. Variations in cancer incidence and mortality exist by race/ethnicity, sex, age, and socioeconomic status. While increased attention is being given to describing cancer-related disparities, the factors that give rise to these disparities, and how they are interrelated, is poorly understood. This report presents Delaware cancer rates by race/ethnicity to determine the extent of these disparities. The rate of difference in the incidence of and mortality from female breast, colorectal, lung and bronchus, and prostate cancer, and these four major cancers combined among different racial/ethnic groups in Delaware and how these rates compare with those of the United States as a whole is described. The effect of sex, age, and county of residence on disparities in cancer incidence and mortality is also examined, and potential factors that may explain the variations in incidence and mortality by race/ethnicity are presented. Specifically, we include a description of population-based modifiable risk factors and screening measures available. Because stage of disease at diagnosis is a major prognostic factor, one that potentially contributes to differential mortality rates, this report describes variations in the stage of disease for the major cancers. Cancer treatments received are examined and compared across racial groups. 1.1. DELAWARE CANCER STATISTICS Twenty-five percent of all deaths in Delaware from 1999 to 2002 were caused by cancer (1). The average annual age-adjusted cancer mortality rate from 1998 to 2002 was 211.6 per 100,000 persons in Delaware, compared with 197.9 among all states (2). The American Cancer Society (ACS) estimated that in Delaware in 2005, 3,800 cancer cases would be diagnosed and 1,580 deaths would be attributed to cancer (3). The most commonly diagnosed cancers from 1998 to 2002 were lung and bronchus (16 percent), female breast (15 percent), prostate (15 percent), and colorectal cancer (12 percent). These four cancers are also among the leading causes of cancer deaths in the State (2). 1.2. DISPARITIES IN CANCER BURDEN Research conducted across the United States has observed that the burden of cancer varies by race/ethnicity and socioeconomic status (4–8). According to the most recently released United States data from 1997 to 2001 for all cancer sites combined, African American men are 20 percent more likely to be diagnosed with cancer than White men, while African American women are 10 percent less likely to be diagnosed with cancer than White women. African American men are 60 percent more likely to be diagnosed with prostate cancer, 50 percent more likely to be diagnosed with lung and bronchus cancer, and 10 percent more likely to be diagnosed with colorectal cancer than White men. Among women, African American women are 20 percent more likely to be diagnosed with colorectal cancer, equally likely to be diagnosed with lung and bronchus cancer, and 20 percent less likely to be diagnosed with breast cancer than White women (9). There is further evidence of disparity by race/ethnicity in mortality rates. Examining overall data from the United States, African American men are 40 percent more likely to die of cancer than White men, and African American women are 20 percent more likely to die of cancer than White women. The largest disparity is observed in prostate cancer mortality rates, which are 2.4 times higher for African American men than White men. Increased mortality in African Americans is also observed for colorectal, female breast, and lung and bronchus cancer: five-year survival rates are higher for Whites than for African Americans, even after adjusting for differences in the distribution of tumor stage (9). 1.2.1. Factors That Contribute to Disparities Differences that exist both within and between racial/ethnic groups can be described by many factors. A recent report by the Institute of Medicine proposed that health disparities arise as a result of a complex relationship between social, economic, and cultural factors (10, 11). Potential explanations attribute these racial/ethnic disparities in cancer incidence and mortality to dissimilarities in 1) exposure to cancer risk factors, including unhealthy diets and cancer-causing agents, 2) socioeconomic status, and 3) access to early detection services and quality medical care (10–12). To address and ultimately eliminate health disparities, Congress passed the Minority Health and Health Disparities Research and Education Act of 2000, which created the National Center on Minority Health and Health Disparities (NCMHD) as a department within the National Institutes of Health (NIH). In addition, the Center to Reduce Cancer Health Disparities was created within the National Cancer Institute (NCI) to focus specifically on the issue of health disparities in cancer. The goal of both these centers is to increase research on health disparities, training for careers in health disparities, and community intervention and information dissemination to minimize health disparities. In the NCMHD strategic plan, the model outlined in figure 1 is used to describe the complex interaction of factors that are thought to create health disparities (13). The items highlighted in red are aspects that are considered in this report. Figure 1. A model for the development of health disparities National data suggest that there is evidence of disparities in behavioral risk factors. The national Behavioral Risk Factor Surveillance System (BRFSS) data suggest that minorities are less likely to exercise and more likely to be overweight or obese as compared with Whites (14). Research has examined the degree to which socioeconomic status affects disparities in cancer incidence and mortality rates among racial/ethnic groups. Studies that examined cancer diagnosis and mortality have found that socioeconomic status is a stronger predictor than race/ethnicity (8). Socioeconomic status, particularly in terms of income and education, is a strong determinant of working in high-risk occupations and of access to health care; these factors are associated with both developing and surviving cancer. Disproportionate numbers of racial/ethnic minorities compared with Whites live at or below the poverty level. African Americans are twice as likely and Hispanics are four times more likely to have less than a high school education than Whites. Income generally increases with educational attainment; however, at each level of education, income is typically higher for Whites and Asians/Pacific Islanders than African Americans and Hispanics. According to a recent report on the health status of the Nation, men aged 25–64 with annual incomes of less than $10,000 were 2.4 times more likely to die of lung and bronchus cancer than those with annual incomes of $25,000 or more (15). Less-educated men and women had mortality rates two to three times higher than individuals with education levels beyond high school (16, 17). Use of screening tests is strongly associated with health insurance and a reliable source of care (18). National data have shown that, overall, use of screening tests (including female breast examinations and mammography and colorectal and prostate examinations) among African Americans is comparable with their use among Whites. However, in the United States 11 percent of the African American population and almost 23 percent of the Hispanic population report no health insurance or reliable source of care, compared with fewer than seven percent of Whites (19). 1.3. OBJECTIVES An analysis of cancer incidence and mortality was conducted to measure the racial/ethnic disparities in these outcomes in Delaware. Potential determinants of these disparities were also analyzed, including aspects of primary (behavioral risk factors), secondary (screening), and tertiary (treatment) prevention, and previous literature on societal and policy factors was explored. Specific objectives of this report include: * Describing differences in incidence and mortality rates for site-specific cancers and all cancer sites combined among racial/ethnic groups in Delaware by age, sex (where applicable), and county of residence * Examining changes in cancer incidence and mortality in Delaware over time by race/ethnicity, and sex * Exploring whether there are disparities in access to health care, modifiable behavioral risk factors, use of screening tests, stage at diagnosis, and cancer treatments received The literature on cancer-related community knowledge, beliefs, and behaviors; patient and provider factors; and service gaps, policy, and societal factors that may contribute to our understanding of cancer incidence and mortality disparities are briefly summarized. 2. METHODS The methods section is divided by objective. Within each objective, the data sources are described, created variables are explained, and the analytic methods are outlined. This structure was chosen because each data source had a different subset of information available, and the objective drove the choice of data source and variables. 2.1. INCIDENCE AND MORTALITY DATA 2.1.1. Data Sources Data used in the analyses derived from both state and federal sources. 2.1.1.1 Delaware Cancer Registry Delaware cancer incidence data were obtained from the Delaware Cancer Registry (DCR), the State’s central cancer information center and part of the Health Information and Epidemiology section of the Delaware Department of Public Health. Delaware is one of 45 States supported by the National Program of Cancer Registries (NPCR) of the Centers for Disease Control and Prevention (CDC). DCR is population-based, collecting data on all cancer patients who are residents of Delaware at the time of diagnosis. DCR collects information on newly diagnosed cancer cases, cancer treatments received, and cancer deaths, as well as follow-up data. As stated in the Delaware Cancer Control Act of 1980, the purpose of the registry is to “ensure an accurate and continuing source of data concerning cancer and certain specified tumors of a benign nature.” The confidentiality of patient information in the registry database is a requirement of this law (20). Quality assurance is performed by a certified tumor registrar to ensure that the registry database includes complete and accurate data that conform to standards established by the North American Association of Central Cancer Registries (NAACCR) and NPCR. DCR’s 1997, 1998, 1999, and 2002 incidence data have been certified by NAACCR as meeting standards for high-quality data. Table A1 in appendix A lists the criteria for and results of the NAACCR certification for Delaware’s 2002 incidence data. DCR also submits data annually to NPCR. DCR’s data submission in 2005 of cases diagnosed from 1998 to 2002 met NPCR’s standards for quality, completeness, and timeliness. 2.1.1.2 Surveillance, Epidemiology, and End Results Program The national cancer incidence data used for comparison with the Delaware cancer incidence data were provided by the NCI Surveillance, Epidemiology, and End Results (SEER) Program. SEER was established in support of the National Cancer Act of 1971, which mandates the collection, analysis, and dissemination of data used in the prevention, treatment, and analysis of cancer. Data from five States, six metropolitan areas, and the Alaska Native registries are used in the national analyses. These areas include Connecticut, Hawaii, Iowa, New Mexico, Utah, Atlanta, Detroit, Los Angeles, San Francisco/Oakland, San Jose/Monterey, and Seattle/Puget Sound and cover approximately 14 percent of the population of the United States. SEER data are currently the most accurate source of cancer incidence data and therefore were used as a proxy for disparities observed in the United States. Table 1 compares some of the demographics for Delaware and the SEER population. Information on SEER characteristics were provided by NCI (21), and Delaware data were collected from the U.S. Census Bureau (22). Table 1. Comparison of Delaware with the SEER Population Race/Ethnicity Delaware SEER Whites African Americans American Indians/Alaska Natives Asians/Pacific Islanders Hispanics 77 20 0.4 2.2 4.8 67 11 1.4 7.4 19 Live below poverty level 6.5 12 High school graduate 83 78 Live in urban areas 80 89 Foreign born 5.7 15 Delaware differs from the SEER population with regard to the factors examined above; this will have considered when comparing Delaware statistics with SEER data. 2.1.1.3 National Center for Health Statistics National mortality data were obtained from the National Center for Health Statistics (NCHS) and were accessed using SEER*Stat. Cancer deaths from 1980 to 1998 were coded in International Classification of Diseases, Ninth Revision (ICD-9) format, and deaths that occurred between 1999 and 2002 were coded using the ICD, Tenth Revision (ICD-10). 2.1.1.4 Census Data Population estimates by race/ethnicity, sex, and age for the United States and for Delaware and its three counties (Kent, New Castle, and Sussex) were obtained from NCI’s datasets based on data supplied by the U.S. Census Bureau for 1998–2002 (23). All rates were then age-adjusted to the U.S. population using the 2000 census standard distribution. Other cancer reports in Delaware use the Delaware Population Consortium population estimates (24), which may lead to small differences between results included in this report and other reports. 2.1.2 Variable Definitions Cancer incidence and mortality statistics are provided for all cancer sites combined and individually for female breast, colorectal, lung and bronchus, and prostate cancer. For each cancer site, statistics are provided by race/ethnicity, age, sex, and county of residence. The category of all cancer sites was defined by: * Restricting to cases with valid data for age, sex, race/ethnicity, and year of diagnosis * Restricting to malignant cancers, except for urinary bladder cancer where in situ is included * Excluding non-melanoma skin cancer * Excluding prostate cancers with a histology code of 8148 Individual cancers were defined using ICD for Oncology, Third Edition (ICD-O-3) codes for malignant cancers (25). Colorectal cancer was defined using codes C18.0–18.9, C26.0, and C20.9; lung and bronchus cancers were defined using codes C34.0–34.9; female breast cancer was defined using codes C50.0–C50.9, excluding 50.7; and prostate cancer was defined using code 61.9 (except for cases with an ICD-O-3 histology code of 8148). These definitions and exclusions are similar to those used by NPCR, and the guidelines can be found in the NPCR data submission guidelines (26). All criteria, except the exclusion for prostate cases with a histology code of 8148, were in place during the entire five-year period between 1998 and 2002. The exclusion for prostate cancer came into effect only in 2001, but we excluded it throughout for consistency. This histologic code was rare and lead to the exclusion of only two cases. There was minimal impact from restricting our analyses to individuals with valid demographic data. Information on individuals missing race/ethnicity information is presented in table A2 in appendix A. Race/ethnicity was categorized as White, African American, Hispanic, Asian/Pacific Islander, and American Indian/Alaska Native. Individuals of any race with a Hispanic ethnicity were categorized as Hispanic. Age was grouped as follows: 20–39 years, 40–49 years, 50–64 years, 65–79 years, and 80 or older. 2.3 Analytic Methods 2.3.1 Direct Standardization Five-year average age-adjusted incidence and mortality rates in Delaware and its counties were computed using data from DCR and NCHS. To measure whether there are disparities in cancer incidence and mortality among racial and ethnic minorities, incidence and mortality rates were compiled for all cancer sites combined and four site-specific cancers (female breast, colorectal, lung and bronchus, and prostate cancer). These cancer sites were chosen because they contribute 56 percent of new cancer cases and 52 percent of cancer deaths and were therefore the most likely candidates to provide a sufficient sample size to address our research objectives. Incidence and mortality rates (per 100,000 population) were calculated as the number of new cancer cases or cancer deaths, respectively, divided by the population counts for subgroups categorized by race/ethnicity, sex, age, and county of residence. Data were combined to compute five-year averages because incidence and mortality case counts for any given year are very small for some subgroups. All rates for State- and county-level analyses were age-standardized by the direct method to the year 2000 standard U.S. population using SEER*Stat (19). We calculated a 95-percent confidence interval around each rate that was computed. Cancer incidence and mortality rates were also calculated using the SEER population as a proxy for national data. Comparisons between Delaware and the SEER data were made by examining rates and 95-percent confidence intervals. (Rates were determined to be comparable if the confidence intervals overlapped.) 2.3.2 Disparities in Incidence and Mortality Disparities in cancer rates were measured by comparing age-, sex-, and county-specific rates for the different racial/ethnic groups. In order to estimate excess risk, rate ratios and rate differences were calculated using Whites as the reference category. The rate ratio was computed by dividing the age-adjusted incidence or mortality rate for any single minority group by the rate for the reference category. When the rate ratio was equal to one, there was no disparity in the rates being compared. If the rate ratio was greater than one, then cancer incidence or mortality was greater in the minority group than among Whites. If the rate ratio was less than one, then cancer incidence or mortality was greater among Whites than in the minority group. A Taylor series expansion was used to generate the 95-percent confidence intervals for rate ratios comparing minorities with Whites (27), using the software application EpiBasic (28). If the confidence intervals for the two rate ratios did not overlap, the ratios were said to be significantly different from one another. In addition to disparity rate ratios, disparity rate differences were calculated. Using Whites as the reference group, the rate for Whites was subtracted from the rate for African Americans. Thus, positive differences indicate that African Americans had a higher incidence or mortality rate, while negative differences imply greater rates among Whites. The 95-percent confidence intervals were calculated using EpiBasic. These two measures are different ways of examining disparity. Rate ratios are used in research to determine causes of an outcome, in this case, racial disparity in cancer incidence and mortality. Rate differences present the absolute difference in the rates between, in this case, African Americans and Whites. Rate differences are more useful for public health research because they allow investigators to understand the scope of and number of people to be impacted by an intervention. A rate ratio of two could indicate a change in a rate from two to four or from 200 to 400, which would have different public health implications. The objective of this analysis is to both understand the causes of racial disparities and inform public health practitioners who will make recommendations for a cancer control plan to eliminate disparities; therefore, we have presented both measures. 2.2.3 Trends in Cancer Incidence and Mortality To examine trends in cancer incidence and mortality from 1980 to 2002, we created graphs with a data point representing the five-year average rate for every five-year increment between 1980 and 2002. Trends were plotted for all cancer sites combined and individually for female breast, colorectal, lung and bronchus, and prostate cancer using data from 1980 to 2002. Data are presented only for Whites and African Americans because other racial and ethnic groups did not have sufficient sample sizes to be included. Trends were examined for both incidence and mortality. We separated trends by men and women since the patterns observed differed by sex. The graphs were examined to determine whether rates were increasing or decreasing over time and whether the differences between rates in African Americans and Whites were increasing or decreasing over time. 2.2.4 Data Reporting Rules In order to maintain the confidentiality of Delaware residents diagnosed with cancer, when frequency data are presented, cells with five or fewer people are not displayed. In addition, rates that are based on 25 or fewer people in the numerator are considered unstable and are not presented. 2.3 BEHAVIORAL CANCER RISK FACTORS AND SCREENING USAGE 2.3.1 Data Sources The BRFSS was created to survey personal health behaviors and accompanying risk factors that influence premature morbidity and mortality at the State and national levels among individuals aged 18 or older. For this analysis, we selected relevant elements from the BRFSS survey in order to assess potential risk factors and screening measures for identified cancer incidence and mortality rate discrepancies in Delaware. The results were analyzed to determine what health behavior and risk factors could be associated with disparities in cancer detection and treatment in Delaware. Indicators included mammography screening (female breast cancer), tobacco use (lung and bronchus cancer), and various health risk behavior indicators such as health status, alcohol use, and obesity (associated with most forms of cancer). The wording of the BRFSS questions used is included in table C1 of appendix C. 2.3.2 Variable Definitions For this analysis, BRFSS data collected in 2002 were used to obtain prevalence estimates of modifiable cancer risk factors, including tobacco use, exercise, diet, obesity, and alcohol consumption. The data for 2002 were used because they included all the variables of interest and matched the last year of data from the cancer registry. For each of the five risk factors, we categorized respondents based on whether they were at risk. Individuals were considered to be at risk if they 1) were current or former smokers, 2) reported no occupational or leisure-time physical activity, 3) ate fewer than five servings of fruits and vegetables a day, 4) had a body mass index (BMI), a ratio between an individual’s height and weight, of 25 or greater, or 5) were chronic drinkers (women who drank two or more drinks per day or men who drank three or more drinks per day). The prevalence of colorectal and prostate cancer screening was estimated among respondents aged 50 years or older. Colorectal cancer screening was estimated using two different tests: 1) a home-administered fecal occult blood test kit within the past year or 2) a sigmoidoscopy or colonoscopy in the past five years. Prostate cancer screening was estimated for men who reported having had a prostate-specific antigen (PSA) test or digital rectal exam in the past year. Prevalence estimates of mammography screening during the past two years were reported for women aged 40 years or older; the prevalence of receiving clinical breast exams in the past two years was reported for all women. It should be noted that for all the screening tests, the BRFSS does not distinguish between tests performed for screening purposes and tests performed for diagnostic purposes. Race was defined as White, African American, Hispanic, and Other. Other includes Asians, Pacific Islanders, American Indians, and Alaska Natives. Individuals of any race who have a Hispanic ethnicity are included in the Hispanic category, not their race category. For behavioral factors, age was grouped as follows: younger than 40 years, 40–49 years, 50–64 years, 65–79 years, and 80 years or older. Education was categorized as less than high school education, a high school graduate with no further education, a high school graduate with one to three years of college, and college graduates. Income referred to the annual household income from all sources and was categorized as less than $25,000, $25,000–49,999, and $50,000 or more. Health care access was measured by having any type of health insurance coverage, a doctor that the individual considered to be his or her personal doctor, and a usual source of care (defined as seeking non-emergency care in physicians’ offices, clinics, or community health centers and not at emergency rooms or urgent care centers). 2.3.3 Analytic Methods Data analyses were conducted using SAS version 9.1 (29). The prevalence of behavioral risk factors for cancer, access to health care, and use of screening among Delaware residents was compared with BRFSS data for the entire United States (30). Data were presented only for groups that had at least 50 people in the denominator. A table with the denominator for all questions by race/ethnicity is included as table C2 in appendix C. This standard is used by CDC for the presentation of BRFSS data (14). Multiracial individuals were excluded from the analysis because of small sample sizes. We were also unable to examine the prevalence of cancer screening tests among Hispanics. There were 120 Hispanics included in the 2002 sample of the BRFSS; however, most of the screening tests are sex-specific, which decreased the number by half. Furthermore, 60 percent of the Hispanics were younger than 40, below the minimum age for screening. All prevalence estimates were weighted to reflect Delaware’s population distribution. Multivariate logistic regression models were used to predict each behavioral risk factor or screening test by the following independent variables: race, age, sex, county of residence, socioeconomic status, and access to health care. Models were also created to predict each of the access-to-health-care variables by race, age, sex, county of residence, and socioeconomic status. Therefore, each model was adjusted for all the independent variables. The logistic regression procedure accounted for the sampling and weighting used in the BRFSS. The available variables to measure socioeconomic status were income and education, but we chose to include only education in the model because data on education were available for 99.9 percent of the population, while data on income were available for only 71.3 percent. Education and income have a correlation coefficient of 0.42; this suggests that they are moderately correlated, making education a reasonable marker for income. 2.4 CANCER STAGE AT DIAGNOSIS 2.4.1 Data Sources The association between stage at diagnosis and variables related to disparities was examined using DCR data. Data were aggregated from 1998 to 2002, the five most recent years of available data. Disparities in the stage at diagnosis for female breast, colorectal, and prostate cancer were examined because these cancers have recommended screening tests. 2.4.2 Variable Definitions Stage of cancer was categorized using the SEER summary stage, a scale that categorizes cancers as in situ, local, regional, distant, or unstaged, using the following definitions: * In situ—Presence of malignant cells within the cell group from which they arose * Local—Invasive neoplasm confined entirely to the site of origin * Regional—Tumors that have extended beyond the limits of the site of origin * Distant—Tumors that have spread to parts of the body remote from the primary site of origin * Unstaged—Tumors with insufficient information to assign a stage SEER summary stage does not account for tumor size or other pathological features, as does the American Joint Committee on Cancer’s tumor, node, and metastasis classification system; however, SEER summary stage is the most routinely collected and allows for comparison with SEER data. In situ tumors were excluded, but all other stages were examined. 2.4.3 Analytic Methods Disparities in the stage at diagnosis were examined by race/ethnicity, age, sex, and county of residence. Race/ethnicity data were categorized as White, African American, Hispanic, Asian/Pacific Islander, and American Indian/Alaska Native. Hispanics included individuals of all races. Data on American Indians/Alaska Natives did not meet the requirement of at least 50 individuals in the denominator and therefore were not presented; data on Hispanics and Asians/Pacific Islanders were presented in crude tables, but the data were too sparse for stratified analyses. Disparities in stage at diagnosis were examined by age to see whether the disparities were observed to a lesser or greater extent in those who were younger than 40 for female breast cancer and younger than 50 for colorectal and prostate cancer because those individuals were too young for screening to be recommended. Older individuals were also examined to determine whether Medicare coverage eliminated disparities. All newly diagnosed cancers between 1998 and 2002 that were coded as unstaged were removed from the analysis. In table A3 in appendix A, the proportion of cases coded as unstaged are presented for each cancer type and compared with the percentage unstaged in the SEER data. All cancers classified as regional or distant were combined to form a category that represented cancers diagnosed at advanced stages. Patients diagnosed with regional and distant cancers usually have a less favorable prognosis than those diagnosed with local cancer. 2.5 CANCER TREATMENT Initial attempts to examine disparities in cancer treatment used only the treatment-related data resident in the DCR; i.e., did not include review of any source records from which the DCR data derive. The data formats and variable definitions conformed to conventions outlined by the NAACCR version 10.1 data dictionary (31). The analyses were restricted to cancers diagnosed at local disease stage because the treatment guidelines are clearer; table A3 in appendix A shows the proportion of cases that were therefore not included in the treatment analyses. The treatment analyses were restricted to individuals whose cancer was reported from a facility in Delaware and whose cancer was diagnosed or treated – or both diagnosed and treated – at that facility. To compare treatments, we created a variable to represent “standard” treatment in 2001. Standard treatment was defined using the NCI Physician Data Query as reported in an article by Shaver et al. (32). This variable was created for female breast, colorectal, lung and bronchus, and prostate cancer. A description of the definition and codes used to categorize individuals into treatment groups for each of the cancers examined is included in table A4 of appendix A. The risk of not receiving standard treatment was examined using multivariate logistic regression; analyses were performed using SAS version 9.1 (29). Results of these initial analyses strongly suggested DCR treatment data were not sufficiently complete/accurate to support this type of analysis without reference to source documents. To address that limitation, the Delaware Cancer Consortium undertook a validation study to determine if the treatment-related data in the DCR accurately reflected the treatment received, as documented in individual facility patient and cancer registry records. The validation study was limited to breast and colorectal cancer cases, and included all cases (n = 334) designated during the initial analyses as not having received standard treatment. All reviews were conducted by the physician oncologist members of the Consortium, and were designed to capture both treatments received and justifiable reasons for a lack of treatment, e.g., patient refusal or the presence of co- morbid conditions precluding certain treatments. Both patients whose source documents reflected receipt of standard treatment and those whose lack of standard treatment was deemed justifiable were designated as having received “appropriate” treatment. Review results were incorporated into all subsequent treatment analyses. 2.6 DATA INTERPRETATION Data in this report are presented as rates, rate ratios, and odds ratios, each with a corresponding 95-percent confidence interval. In addition, percentages and percentage differences are displayed. There are a number of ways that the data can be interpreted. From a clinical or public health standpoint, the definition of an important result is a difference that exceeds what is considered clinically acceptable or what we have the knowledge and resources to address. Statistically significant results are defined as rate ratios and odds ratios where the confidence interval does not include one. In situations where there is a comparison of two rates, rate ratios, or odds ratios, statistically significant results are situations where the confidence intervals do not overlap. To give readers the ability to identify statistically significant results, two measures are indicated as being different only if they are statistically significantly different. Results that are potentially important but not statistically significant are noted; however, in these cases it is stated that the data suggest there may be an effect. 3. RESULTS 3.1. CANCER INCIDENCE AND MORTALITY The objective of this section is to determine whether there are racial/ethnic disparities in cancer incidence and mortality, and in situations where disparities exist to measure the magnitude of the disparity. The results of the analyses using incidence data from DCR and mortality data from NCHS from 1998 to 2002 are reported, and the contribution of race/ethnicity, age, sex, and county of residence to the disparities in cancer incidence and mortality are discussed. The results are presented for all cancer sites combined and by site for female breast, colorectal, lung and bronchus, and prostate cancer. The results are first presented for racial/ethnic disparities, and then, stratifying by race, we present the results for sex, age, and geographic disparities. 3.1.1. Cancer Burden in Delaware Table 2 presents the number of Delaware residents who were diagnosed with and who died from cancer between 1998 and 2002 for all cancer sites combined and for each of the four major cancer sites by race/ethnicity. Table 2. Number of New Cancers and Deaths from Cancer in Delaware, by Race/Ethnicity Incidence Mortality White African American Hispanic Asian White African American Hispanic Asian All sites 16,416 2,867 187 150 7,002 1,249 80 42 Female breast 2,322 392 23 31 513 115 6 < 6 Colorectal 1,841 324 9 19 676 140 < 6 < 6 Lung and bronchus 2,609 418 18 13 2,128 335 18 8 Prostate 2,448 591 24 13 328 102 < 6 < 6 Sources: DCR; NCHS, 1998–2002. 3.1.2. Disparities in Cancer Incidence and Mortality The Delaware and U.S. age-adjusted incidence and mortality rates from 1998 to 2002 for all cancer sites combined and by site for female breast, colorectal, lung and bronchus, and prostate cancer are presented in this section. Figures 2 and 3 show the incidence and mortality rates for all cancer sites combined for Whites, African Americans, Hispanics, and Asians/Pacific Islanders in Delaware and the United States. Figures 4 and 5 display the incidence and mortality rates individually for female breast, colorectal, lung and bronchus, and prostate cancer. For the remainder of the section, we present data only on Whites and African Americans because the data were too sparse for Hispanics and Asians/Pacific Islanders. Table 3 presents the incidence and mortality rate ratios comparing African Americans with Whites for Delaware and the United States. For further detail, the numbers and age-adjusted incidence and mortality rates with 95-percent confidence intervals by race are displayed in appendix B. In section 3.1.2., figures displaying incidence rates use a y-axis ranging from 0/100,000 to 600/100,000, while figures displaying mortality rates range from 0/100,000 to 300/100,000. Figure 2. Age-adjusted incidence rates for all cancer sites combined, Delaware and the United States Sources: DCR; SEER Program, 1998–2002. * In Delaware, African Americans had a higher cancer incidence than Whites for all cancer sites combined, and rates for Hispanics and Asians/Pacific Islanders were lower than for Whites. * Among Whites, cancer incidence was similar in Delaware and the United States, but the rate in African Americans was higher in Delaware than in the United States, though not significantly. Cancer incidence rates in Hispanics and Asians/Pacific Islanders were lower in Delaware than in the United States. Figure 3. Age-adjusted mortality rates for all cancer sites combined, Delaware and the United States Source: NCHS, 1998–2002. * Cancer mortality was higher among African Americans than Whites in Delaware, while rates among Hispanics and Asians/Pacific Islanders were lower than for Whites. * Mortality rates in Delaware were higher than in the United States for all races/ethnicities except Asians/Pacific Islanders but only significantly higher among Whites. Figure 4. Age-adjusted incidence rates for the four major cancer sites, by race, Delaware and the United States Sources: DCR; SEER Program, 1998–2002. * In Delaware, African Americans had a significantly higher incidence than Whites for colorectal and prostate cancer. Incidence rates for lung and bronchus cancer and for female breast cancer did not differ significantly between African Americans and Whites in Delaware. * No differences were observed between Delaware and the United States for African Americans; i.e., incidence rates for African American Delawareans are similar to those for African Americans throughout the United States.. * Compared with the United States, rates of female breast cancer and of prostate cancer were significantly lower in Delaware among Whites, while rates of lung and bronchus cancer were significantly higher among Whites. Figure 5. Age-adjusted mortality rates for the four major cancer sites, by race, Delaware and the United States Source: NCHS, 1998–2002. * Mortality was higher in Delaware among African Americans for all the major cancer sites - even for female breast cancer, where incidence was lower. The differences were significant for all sites except lung and bronchus cancer. * For both Whites and African Americans, mortality rates were higher in Delaware than in the United States for most of the major cancer sites; mortality rates were lower in Delaware than in the United States for prostate cancer. Table 3. Incidence and Mortality Rate Ratios (With 95-Percent Confidence Intervals) Comparing African Americans With Whites in Delaware and the United States Cancer Site Incidence Rate Ratios Mortality Rate Ratios Delaware United States Delaware United States All sites 1.08 (1.04, 1.13) 1.06 (1.05, 1.07) 1.21 (1.14, 1.29) 1.27 (1.27, 1.28) Female breast 0.90 (0.81, 1.01) 0.83 (0.81, 0.85) 1.33 (1.09, 1.63) 1.34 (1.32, 1.36) Colorectal 1.19 (1.06, 1.34) 1.17 (1.14, 1.19) 1.47 (1.22, 1.76) 1.40 (1.38, 1.41) Lung and bronchus 1.06 (0.96, 1.18) 1.26 (1.23, 1.28) 1.08 (0.96, 1.21) 1.15 (1.14, 1.16) Prostate 1.68 (1.53, 1.84) 1.60 (1.56, 1.63) 2.48 (1.98, 3.09) 2.46 (2.43, 2.49) Sources: DCR; SEER Program; NCHS, 1998–2002. The incidence rate ratio for Hispanics in Delaware was 0.51 (0.44, 0.59), and for Asians the ratio was 0.58 (0.49, 0.68). The mortality rate ratio for Hispanics in Delaware was 0.72 (0.58, 0.90), and for Asians it was 0.50 (0.37, 0.68). The incidence rate ratios show that: * Cancer incidence in Delaware among African Americans was 8 percent greater for all sites, 19 percent greater for colorectal cancer, and 68 percent greater for prostate cancer. The data suggest that incidence among African Americans was lower for female breast cancer and slightly elevated for lung and bronchus cancer. * Rates among Hispanics and Asians were lower than for Whites. * Disparities in lung and bronchus cancer incidence were lower in Delaware than in the United States; disparities for other cancer sites were comparable. The mortality rate ratios show that: * Cancer mortality in Delaware among African Americans was 21 percent higher for all sites, 33 percent higher for female breast cancer, 47 percent higher for colorectal cancer, and 148 percent higher for prostate cancer. * Cancer mortality was lower among Hispanics and Asians than Whites. * Disparities in Delaware were comparable to the disparities observed in the United States. 3.1.3. Cancer Disparities by Sex, Age, and County of Residence Next, the racial disparities in incidence and mortality were examined to determine whether they differed between males and females, by age at diagnosis, and by whether the individual resided in New Castle, Kent, or Sussex County at the time of his or her diagnosis. The number of new cases and deaths and the incidence and mortality rates with 95-percent confidence intervals for all cancer sites combined and the four major cancer sites separately are included in appendix A. Incidence and mortality rate ratios and rate differences are presented for sex, age, and county of residence. 3.1.3.1. Disparities by Sex In table 4 and in figures 6 and 7, the incidence and mortality rate ratios and rate differences are presented. Data examining disparities between males and females are presented for all cancer sites, colorectal cancer, and lung and bronchus cancer. The figures displaying rate differences in this section use a y-axis that ranges from –15/100,000 to 45/100,000. Table 4. Incidence and Mortality Rate Ratios (With 95-Percent Confidence Intervals) Comparing African Americans With Whites, by Sex, in Delaware and the United States Cancer Site Incidence Rate Ratios Mortality Rate Ratios Delaware United States Delaware United States All sites Male Female 1.19 (1.13, 1.26) 0.98 (0.93, 1.04) 1.22 (1.20, 1.23) 0.93 (0.92, 0.94) 1.30 (1.19, 1.41) 1.17 (1.07, 1.27) 1.40 (1.39, 1.41) 1.18 (1.17, 1.19) Colorectal Male Female 1.12 (0.94, 1.33) 1.27 (1.08, 1.49) 1.16 (1.12, 1.20) 1.21 (1.17, 1.26) 1.37 (1.05, 1.79) 1.58 (1.23, 2.03) 1.40 (1.38, 1.42) 1.43 (1.41, 1.46) Lung and bronchus Male Female 1.11 (0.97, 1.28) 1.09 (0.93, 1.27) 1.44 (1.40, 1.49) 1.08 (1.04, 1.12) 1.18 (1.02, 1.37) 0.99 (0.83, 1.19) 1.35 (1.33, 1.36) 0.95 (0.94, 0.97) Sources: DCR; SEER Program; NCHS, 1998–2002. * In Delaware, the data suggest there was greater disparity in cancer incidence for all sites combined among men than among women, and greater disparity in colorectal cancer incidence among women than among men. There was no difference by sex for lung and bronchus cancer. * There was greater disparity among men with lung and bronchus cancer in the United States than in Delaware. For other sites, disparities by sex for cancer incidence were comparable. * The data suggest that disparity in cancer mortality in Delaware was greater among men for all cancer sites combined and lung and bronchus cancer and greater among women for colorectal cancer. * Disparities in cancer mortality by sex were comparable in Delaware and the United States. Figure 6. Age-adjusted incidence rate differences comparing African Americans with Whites for colorectal and lung and bronchus cancer, by sex, Delaware and the United States Sources: DCR; SEER Program, 1998–2002. * Among women in Delaware there were significant disparities between African Americans and Whites for colorectal cancer. There was no evidence of a difference in disparities between men and women in Delaware and in the United States. * A difference in disparities in lung and bronchus cancer incidence were observed in the United States, where men had a higher incidence than women, but not in Delaware. Figure 7. Age-adjusted mortality rate differences comparing African Americans with Whites for colorectal and lung and bronchus cancer, by sex, Delaware and the United States Source: NCHS, 1998–2002. * The data suggest that African American women with colorectal cancer were more likely to die than White women, and African American men were more likely to die of lung and bronchus cancer than White men. There was no difference between men and women for lung and bronchus cancer, but the data suggest that disparities were greater among men with lung and bronchus cancer. * Disparities in the United States were comparable to disparities in Delaware, but the data suggest greater disparity for lung and bronchus cancer among men in the United States than among men in Delaware. 3.1.3.2. Disparities by Age In table 5 and in figures 8 and 9, the incidence and mortality rate ratios and differences are presented by age at diagnosis. We examined age at diagnosis using age groupings that considered ages for screening recommendations (40+ for female breast cancer and 50+ for colorectal and prostate cancer) and Medicare eligibility (65+), since these factors could be related to access to screening and health care, which may be a cause of cancer disparities. Data are presented for all cancer sites combined and for the four major cancer sites individually. Figures that display the rate differences use graphs with a y-axis that ranges from –400/100,000 to 800/100,000. Table 5. Incidence and Mortality Rate Ratios (With 95-Percent Confidence Intervals) Comparing African Americans With Whites, by Age, in Delaware Age (Years) Incidence Rate Ratios Mortality Rate Ratios All sites 20–39 40–49 50–64 65–79 80+ 0.86 (0.75, 1.00) 1.17 (1.04, 1.31) 1.15 (1.08, 1.24) 1.01 (0.95, 1.08) 1.06 (0.94, 1.20) 0.99 (0.71, 1.38) 1.53 (1.23, 1.90) 1.25 (1.11, 1.40) 1.18 (1.07, 1.29) 1.12 (0.98, 1.29) Female breast 20–39 40–49 50–64 65–79 80+ 1.36 (0.96, 1.92) 1.24 (1.00, 1.53) 0.81 (0.67, 0.98) 0.70 (0.55, 0.88) 0.94 (0.65, 1.35) NA 2.41 (1.46, 3.98) 1.78 (1.24, 2.53) 1.06 (0.72, 1.57) NA Colorectal 20–39 40–49 50–64 65–79 80+ NA 1.43 (0.96, 2.12) 1.18 (0.94, 1.49) 1.06 (0.88, 1.28) 1.27 (0.98, 1.65) NA NA 1.63 (1.13, 2.35) 1.38 (1.05, 1.82) 1.35 (0.93, 1.96) Lung and bronchus 20–39 40–49 50–64 65–79 80+ NA 1.49 (1.04, 2.14) 1.20 (1.01, 1.42) 0.94 (0.80, 1.10) 0.97 (0.70, 1.35) NA 1.52 (0.96, 2.40) 1.15 (0.94, 1.41) 1.01 (0.85, 1.19) 1.00 (0.73, 1.36) Prostate 20–39 40–49 50–64 65–79 80+ NA NA 1.94 (1.69, 2.23) 1.50 (1.31, 1.72) 1.42 (1.02. 1.97) NA NA NA 3.02 (2.22, 4.11) 2.15 (1.50, 3.09) NA = Rates based on counts too small to be displayed. Sources: DCR; NCHS, 1998–2002. * For both incidence and mortality there was no statistically significant evidence that disparity differed by age at diagnosis or death. * However, the data suggest some patterns with respect to age: * While sufficient data at young ages were not always available, the available data showed that the disparities were greater at younger ages. * Disparities among individuals aged 65 or older were usually less than among younger individuals. Figure 8. Incidence rate differences comparing African Americans with Whites for the four major cancers, by age, in Delaware Source: DCR, 1998–2002. * There was no evidence of significant differences in cancer disparity by age on the absolute scale. Figure 9. Mortality rate differences comparing African Americans with Whites for the four major cancers, by age, in Delaware Source: NCHS, 1998–2002. * The data for colorectal and prostate cancer suggest that disparities in cancer mortality increased with age. Minimal disparity was observed for lung and bronchus cancer. 3.1.3.3. Disparities by County of Residence In table 6 and in figures 10 and 11, the incidence and mortality rate ratios and differences are presented by county of residence at diagnosis. We examined county of residence for all cancer sites combined and for the four major cancers by site. Figures that display the rate differences use graphs with a y-axis that ranges from –100/100,000 to 200/100,000. Table 6. Incidence and Mortality Rate Ratios (With 95-Percent Confidence Intervals) Comparing African Americans With Whites, by County of Residence, in Delaware Cancer Site Incidence Rate Ratios Mortality Rate Ratios All sites Kent New Castle Sussex 0.99 (0.90, 1.10) 1.11 (1.06, 1.17) 1.04 (0.95, 1.15) 1.14 (0.99, 1.32) 1.21 (1.12, 1.30) 1.24 (1.08, 1.42) Female breast Kent New Castle Sussex 0.79 (0.59, 1.05) 0.96 (0.85, 1.09) 0.80 (0.60, 1.06) NA 1.31 (1.02, 1.68) 1.84 (1.20, 2.83) Colorectal Kent New Castle Sussex 1.30 (1.00, 1.69) 1.22 (1.01, 1.40) 0.94 (0.69, 1.27) 1.86 (1.20, 2.90) 1.40 (1.11, 1.77) 1.51 (1.01, 2.24) Lung and bronchus Kent New Castle Sussex 0.78 (0.59, 1.02) 1.17 (1.03, 1.33) 1.01 (0.80, 1.28) 0.77 (0.58, 1.03) 1.15 (1.00, 1.33) 1.12 (0.87, 1.45) Prostate Kent New Castle Sussex 1.67 (1.32, 2.11) 1.63 (1.46, 1.82) 1.71 (1.37, 2.14) NA 1.94 (1.42, 2.66) NA NA = Rates based on counts too small to be displayed. Sources: DCR; NCHS, 1998–2002. * For all sites combined and for the specific sites of interest, there was minimal difference in disparities in cancer incidence and mortality by county of residence. * For mortality, data were sometimes too sparse to present for all counties. However, the data suggest that colorectal cancer mortality was higher in Kent and Sussex Counties than in New Castle County, and that lung and bronchus cancer mortality was lowest in Kent County. Figure 10. Incidence rate differences comparing African Americans with Whites for the four major cancers, by county of residence, in Delaware Source: DCR, 1998–2002. * This figure also presents the differences in cancer incidence among African Americans compared with incidence among Whites, by county. Figure 11. Mortality rate differences comparing African Americans with Whites for the four major cancers, by county of residence, in Delaware Source: NCHS, 1998–2002. * The pattern for disparities in cancer mortality by county differed depending on the cancer site, but no significant interaction by county of residence was observed. 3.2. TRENDS IN CANCER INCIDENCE AND MORTALITY In this section, Delaware incidence and mortality rates from 1980 to 2002 are presented for all cancer sites combined and for the four major cancer sites. The analysis was restricted to African Americans and Whites because we did not have sufficient data for the other races. Results for all cancer sites combined, colorectal cancer, and lung and bronchus cancer are presented by sex, since this was an important determinant of the trend. Figures 12–16 include the trends for incidence, and figures 17–21 display the trends for mortality. Since the annual number for some cancers and subgroups of interest were small, the five-year average incidence and mortality rates were plotted for each five-year increment from 1980–1984 through 1998–2002. Trends were examined to determine whether advances in cancer prevention behaviors, screening, and treatment have had a similar impact on African Americans and Whites. We examined these trends to determine whether the rates were increasing or decreasing over time and whether the difference in rates between African Americans and Whites was increasing or decreasing. No statistics are presented for the trend analysis; statements about the data reflect visual inspection. 3.2.1. Cancer Incidence in Delaware, 1980–2002 3.2.1.1. All Cancer Sites Combined Figure 12. Trends in cancer incidence for all sites combined comparing Whites with African Americans, by sex, in Delaware Note: A data point is included for each of the 20 five-year increments, but due to space limitations only every third point is labeled on the x-axis. Source: DCR, 1980–2002. * Cancer incidence rates among White men were consistently lower than those of African American men and started to decline earlier: the incidence rate for White men peaked between 1991 and 1995 at 640/100,000, and the rate for African American men peaked between 1992 and 1996 at 878/100,000. * The five-year average rate for African American men began declining sharply in 1993–1997. * Little difference existed between White and African American women, whose incidence rates were lower than for men and remained stable over time. 3.2.1.2. Female Breast Cancer Figure 13. Trends in female breast cancer incidence comparing Whites with African Americans, in Delaware Note: A data point is included for each of the 20 five-year increments, but due to space limitations only every third point is labeled on the x-axis. Source: DCR, 1980–2002. * Breast cancer incidence was consistently higher in White women than African American women. * There were sharp increases in incidence seen in all women in 1983-1987 and 1987-1991. In recent years, incidence decreased for both White and African American women. 3.2.1.3. Colorectal Cancer Figure 14. Trends in colorectal cancer incidence comparing Whites with African Americans, by sex, in Delaware Note: A data point is included for each of the 20 five-year increments, but due to space limitations only every third point is labeled on the x-axis. Source: DCR, 1980–2002. * Among men, colorectal cancer incidence rates for African Americans were initially lower than for Whites, but due to a steady upward trend, the rates for African Americans surpassed the rates for White men in 1988-1992 but then began to decline, while the rates for White men declined steadily throughout the entire time period. * A similar pattern was seen for women. African Americans began with lower rates, then their incidence increased, peaked in 1985-1989, and then declined. However, beginning in 1993- 1997 their incidence rates have been increasing. * The net result is that disparities in the colorectal cancer incidence rate are widening for women and possibly for men. 3.2.1.4. Lung and Bronchus Cancer Figure 15. Trends in lung and bronchus cancer incidence comparing Whites with African Americans, by sex, in Delaware Note: A data point is included for each of the 20 five-year increments, but due to space limitations only every third point is labeled on the x-axis. Source: DCR, 1980–2002. * Rates among African American men were higher than among White men. Incidence for African American men peaked in 1989-1993 and has been declining steadily since; it is now approaching the rate observed in White men. * Lung and bronchus cancer incidence is lower among women than men; however, rates were increasing until the mid-1990s, but are now decreasing. Rates among African American women were higher than among Whites, but in recent years their incidence has been similar. 3.2.1.5. Prostate Cancer Figure 16. Trends in prostate cancer incidence comparing Whites with African Americans, in Delaware Note: A data point is included for each of the 20 five-year increments, but due to space limitations only every third point is labeled on the x-axis. Source: DCR, 1980–2002. * During the 1980s, prostate cancer incidence rates were low and climbed slowly for Whites; for African Americans there was a decline. * During the early 1990s, rates increased sharply in both races but to a greater extent among African Americans. * Rates began to decline in the mid-1990s, but recent data suggest that rates have reached a plateau and are possibly beginning to increase. 3.2.2. Cancer Mortality in Delaware, 1980–2002 3.2.2.1. All Cancer Sites Combined Figure 17. Trends in cancer mortality for all sites combined comparing Whites with African Americans, by sex, in Delaware Note: A data point is included for each of the 20 five-year increments, but due to space limitations only every third point is labeled on the x-axis. Source: NCHS, 1980–2002. * Cancer mortality was higher among men than women, and for both sexes mortality was greater for African Americans than for Whites. * Mortality rates for African American men peaked in 1989–1993 and have been declining steadily since. * For White men and all women, rates declined slightly over time. In White women less than a 10/100,000 difference in mortality was seen over the 20 years examined. 3.2.2.2. Female Breast Cancer Figure 18. Trends in female breast cancer mortality comparing Whites with African Americans, in Delaware Note: A data point is included for each of the 20 five-year increments, but due to space limitations only every third point is labeled on the x-axis. Source: NCHS, 1980–2002. * After 1988-1992, breast cancer mortality consistently decreased among White women. * Mortality for African American women increased until 1986–1990 then decreased to the extent that in 1991–1995 the rate was comparable to that of White women. However, since then the mortality rate of African American women has been higher than that of White women. 3.2.2.3. Colorectal Cancer Figure 19. Trends in colorectal cancer mortality comparing Whites with African Americans, by sex, in Delaware Note: A data point is included for each of the 20 five-year increments, but due to space limitations only every third point is labeled on the x-axis. Source: NCHS, 1980–2002. * Colorectal cancer mortality steadily declined among White men and women. * Except during 1989-1994, rates among African American men decreased until 1994–1998 and are now increasing. * Rates among African American women increased and decreased over time, but there was little net change. 3.2.2.4. Lung and Bronchus Cancer Figure 20. Trends in lung and bronchus cancer mortality comparing Whites with African Americans, by sex, in Delaware Note: A data point is included for each of the 20 five-year increments, but due to space limitations only every third point is labeled on the x-axis. Source: NCHS, 1980–2002. * After 1989-1994, there was a steady decline in the lung and bronchus cancer mortality rate among African American men; it is now approaching the mortality rate of White men. * The rate for African American women was slightly but consistently higher than for White women until the mid-90’s. The mortality rate of African American women is now similar to the mortality rate of White women. 3.2.2.5. Prostate Cancer Figure 21. Trends in prostate cancer mortality comparing Whites with African Americans, in Delaware Note: A data point is included for each of the 20 five-year increments, but due to space limitations only every third point is labeled on the x-axis. Source: NCHS, 1980–2002. * The prostate cancer mortality rate has been much higher among African American men than White men. * Although there has been a steady decline in the prostate cancer mortality rate among White men, the rate for African American men did not decline until after 1992-1996, when it peaked at 105.6/100,000. * With the mortality rate for African American men declining faster than for White men, the disparity in rate has also been decreasing. 3.3. FACTORS THAT CONTRIBUTE TO DISPARITIES This section addresses factors that may contribute to the disparities in cancer incidence and mortality observed in Delaware. Four areas that are known to influence cancer prevention and control were examined: 1) risk factors, 2) screening, 3) stage at diagnosis, and 4) treatment. The following questions were addressed: * Are there differences in the prevalence of modifiable behavioral risk factors for cancer by race/ethnicity? Are these differences affected by age, sex, county of residence, socioeconomic status, and access to health care? * Does the use of screening tests differ among racial groups in Delaware? Does the use of screening among racial groups vary by age, sex, county of diagnosis, socioeconomic status, and access to health care? * Does the distribution of advanced-stage disease differ by race, age, sex, and county of diagnosis? * Does the proportion of cancer patients with local-stage disease who did not receive the standard treatment differ by race, age, sex, county of diagnosis, and insurance status? 3.3.1. Cancer Risk Factors and Screening Usage BRFSS data were used to examine behavioral cancer risk factors and cancer screening usage in Delaware and to compare Delaware with the United States. Table 7 compares demographic characteristics of the population of Delaware with the United States. Table 7. Comparison of Demographic Characteristics Between Delaware and the United States, 2002 Delaware (%) United States (%) Total White African American Hispanic Total White African American Hispanic Race/ethnicity 78 13 5 70 10 14 Age 20–39 40–49 50–64 65–79 80+ 40 21 21 14 3 37 21 22 16 4 51 21 17 9 2 9 23 15 2 1 42 20 21 13 4 37 21 23 15 4 47 21 19 10 3 8 19 15 7 1 Sex Male Female 48 52 47 53 43 57 57 43 48 52 48 52 44 56 49 51 Education < High school High school graduate Some college College graduate 9 33 26 32 7 34 26 34 14 37 29 20 36 29 17 18 13 31 27 29 8 31 28 32 16 35 29 20 34 29 23 15 Income < $25,000 $25,000–$49,000 $50,000+ 21 30 49 17 29 53 32 39 29 45 26 29 29 32 38 23 33 44 43 34 23 53 29 18 Source: BRFSS, 2002. Compared with the United States: * Delaware had a comparable age and sex distribution. * Delaware had a larger African American population and a smaller Hispanic population. * A lower proportion of the Delaware population had less than a high school education. * A higher proportion of the Delaware population had incomes of greater than $50,000. Tables 8–10 display differences by race/ethnicity in the prevalence of access to health care, behavioral risk factors for cancer, and screening usage. A table displaying the denominator for each variable examined by race/ethnicity is included in table C2 in appendix C. Table 8. Access to Health Care in Delaware and the United States, by Race/Ethnicity No Health Insurance No Personal Doctor No Usual Source of Care % (95% CI) % (95% CI) % (95% CI) Delaware White African American Hispanic 7.6 (6.2, 9.0) 12.8 (8.7, 16.9) 21.3 (9.0, 33.7) 10.1 (8.0, 12.1) 16.1 (11.1, 21.2) 28.7 (15.5, 41.9) 5.7 (4.6, 6.8) 12.4 (8.2, 16.6) 14.1 (6.4, 21.9) United States White African American Hispanic 11.1 (10.9, 11.4) 20.3 (19.3, 21.2) 32.3 (31.0, 33.7) 16.8 (16.5, 17.1) 22.5 (21.4, 23.6) 38.2 (36.8, 39.7) 14.9 (14.6, 15.2) 24.8 (23.7, 25.9) 24.6 (23.3, 25.9) Source: BRFSS, 2002. * Delaware residents of any race/ethnicity were more likely to have access to health care compared with residents of the United States. However, due to small numbers, the differences for Hispanics were not always statistically significant. * In Delaware, Hispanics were less likely to have health insurance, a personal doctor, and a usual source of care than Whites; in most instances, the differences were statistically significant. * The data suggest that in Delaware, African Americans were also less likely to have health insurance, a personal doctor, and a usual source of care than Whites; the only statistically significant difference was in having a usual source of care. Figure 22 compares the percentage difference in health care access for minorities versus Whites in Delaware and in the United States. Using the data from table 8, we examined the difference between the percentage observed in Whites and in non-Whites. Larger differences indicate a larger disparity between Whites and non-Whites. Figure 22. Racial/ethnic disparities in access to health care in Delaware and the United States Source: BRFSS, 2002. * For all three measures of access to health care, disparities were observed for African Americans and Hispanics, compared with Whites. * The data suggest that the disparities in access to health care were greater for Hispanics than African Americans. * There were no significant differences in the magnitude of the disparity when comparing Delaware with the United States, but the data suggest that in general the disparity was less in Delaware. Table 9. Comparison of Modifiable Risk Factors for Cancer Between Delaware and the United States, by Race/Ethnicity No Exercise Former Smokers Current Smokers Overweight (BMI: 25–29) Obese (BMI: 30+) Poor Diet (Fewer Than Five Fruits and Vegetables/Day) Chronic Alcohol Intake1 % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) Delaware White African American Hispanic 17.6 (15.9, 19.4) 37.5 (31.3, 43.6) 21.1 (11.7, 30.5) 28.0 (25.9, 30.2) 17.3 (12.8, 21.8) 31.6 (17.4, 45.8) 25.6 (23.1, 28.1) 25.6 (19.7, 31.4) 18.1 (7.0, 29.1) 35.5 (33.0, 37.9) 41.0 (34.7, 47.5) 40.4 (25.8, 55.0) 20.8 (18.7, 22.8) 33.2 (27.3, 39.2) 18.4 (5.5, 31.3) 79.9 (78.0, 81.7) 87.3 (83.7, 91.1) 79.1 (67.6, 90.6) 8.6 (7.0, 10.2) 5.6 (2.4, 8.9) 9.2 (-1.1, 19.5) United States White African American Hispanic 20.3 (20.0, 20.6) 30.6 (29.4, 31.7) 35.9 (34.5, 37.2) 27.1 (26.8, 27.4) 16.4 (15.6, 17.4) 18.1 (17.0, 19.2) 23.4 (23.0, 23.7) 21.8 (20.8, 22.8) 18.6 (17.4, 19.7) 36.8 (36.4, 37.2) 36.1 (34.9, 37.3) 39.4 (38.0, 40.9) 20.5 (20.2, 20.8) 33.2 (32.1, 34.4) 24.0 (22.7, 25.3) 75.7 (75.3, 80.0) 77.0 (76.0, 78.1) 77.0 (75.8, 78.2) 6.3 (6.1, 6.5) 3.7 (3.2, 4.2) 4.3 (3.7, 4.9) 1 Women who drank two or more drinks per day or men who drank three or more drinks per day. Source: BRFSS, 2002. * African Americans were more likely to have no exercise, be obese, and eat fewer than five fruits and vegetables per day than Whites. * The prevalence of behavioral risk factors was comparable in Whites and Hispanics. * African Americans in Delaware were more likely to have a poor diet than African Americans in the United States overall. No other differences were statistically significant, but the prevalence of risk factors seemed to be slightly higher in Delaware. Figure 23 compares the racial/ethnic disparities in cancer risk factors in Delaware with the disparities observed in the rest of the United States. As in figure 22, we examined the difference between the percentage observed in Whites and the percentage observed in non-Whites. Larger differences indicate a greater disparity between Whites and non-Whites. Negative numbers indicate that Whites are more likely to engage in a risk factor than African Americans or Hispanics. We examined the proportion of study respondents who never exercised, were ever smokers, had a BMI greater than 25, ate fewer than five fruits and vegetables per day, and were chronic alcohol drinkers. Figure 23. Racial/ethnic disparities in behavioral cancer risk factors in Delaware and the United States Source: BRFSS, 2002. * The largest disparities for African Americans were in exercise, smoking and obesity. In the case of smoking, African Americans were less likely to have ever smoked than were Whites. * The disparity in exercise and diet was greater in Delaware than in the United States. * Numbers were small for Hispanics, but the data suggest that they had less disparity than African Americans and that the disparity was less in Delaware than the United States. Table 10 shows the differences in screening behavior for colorectal, prostate, and female breast cancer by race. The results show the prevalence of being screened for cancer within the time interval currently recommended by ACS. Table 10. Comparison of the Percentage of Individuals Who Have Not Received Cancer-Screening Tests Within the Recommended Time Interval Between Delaware and the United States, by Race % Not Receiving Test Fecal Occult Blood Test Sigmoidoscopy or Colonoscopy Prostate-Specific Antigen Test Digital Rectal Exam Mammogram (Females) Clinical Breast Exam (Females) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) % (95% CI) Delaware White African American 76.2 (73.5, 78.9) 81.3 (73.0, 89.6) 48.6 (45.4, 51.9) 53.1 (42.5, 63.8) 36.6 (31.5, 41.8) 30.8 (15.4, 46.2) 53.1 (48.7, 57.6) 45.8 (31.5, 60.0) 16.0 (13.5, 18.5) 12.8 (5.9, 19.8) 14.6 (12.3, 16.8) 14.3 (8.4, 20.1) United States White African American 77.5 (77.0, 77.9) 79.6 (77.8, 81.3) 58.7 (58.1, 59.2) 60.4 (58.3, 62.5) 43.6 (42.7, 44.5) 46.5 (43.0, 50.0) 54.5 (53.7, 55.1) 55.9 (53.2, 58.6) 22.8 (22.3, 23.2) 21.9 (20.2, 23.6) 18.2 (17.8, 18.6) 19.7 (18.4, 21.0) Source: BRFSS, 2002. * In both Delaware and the United States, there were no significant differences in the proportion not receiving screening when comparing Whites with African Americans. * Although the difference is not statistically significant, more African Americans than Whites in Delaware received screening for prostate cancer and for female breast cancer. * The data suggest that Delaware residents were more likely to have received screening than did residents of the United States as a whole. Figure 24. Racial disparities in cancer screening test usage in Delaware and the United States Source: BRFSS, 2002. * No significant differences in disparity were observed between Delaware and the United States. Tables 11, 13, and 15 present the odds ratios and 95-percent confidence intervals from logistic regression models examining the association between race/ethnicity and the markers of access to health care, behavioral risk factors, and screening usage, respectively. Tables 12, 14, and 16 present the odds ratios and 95-percent confidence intervals from multivariate models examining the association between race/ethnicity and the measures above but also include in the model other, potentially related variables.. Odds ratios greater than one indicate that the subgroup was more likely not to have (or to have, as applicable) the measure than the reference group; odds ratios less than one indicate that the subgroup was more likely to have (or not to have) the measure. Table 11. Association Between Race/Ethnicity and Access to Health Care in Delaware Odds Ratio (95% Confidence Interval) No Health Insurance No Personal Doctor No Usual Source of Care Race/ethnicity White African American Hispanic Reference 2.0 (1.5, 2.7) 3.1 (1.9, 5.1) Reference 1.4 (1.0, 1.8) 4.0 (2.6, 6.0) Reference 1.9 (1.4, 2.5) 3.5 (2.1, 5.8) Source: BRFSS, 2002. * This table indicates that compared with Whites, African Americans and Hispanics are more likely to lack health insurance, a personal doctor, and a usual source of care. Table 12. Multivariate Regression Model of Predictors of Lack of Access to Health Care No Health Insurance No Personal Doctor No Usual Source of Care Race/ethnicity White African American Hispanic Reference 1.5 (1.1, 2.0) 1.6 (0.93, 2.8) Reference 1.0 (0.76, 1.4) 2.3 (1.4, 3.7) Reference 1.6 (1.2, 2.3) 2.7 (1.5, 4.7) Age < 40 40–49 50–64 65–79 80+ Reference 0.56 (0.41, 0.78) 0.51 (0.37, 0.70) 0.70 (0.04, 1.4) 0.20 (0.09, 0.44) Reference 0.45 (0.33, 0.60) 0.38 (0.28, 0.51) 0.24 (0.16, 0.36) 0.23 (0.11, 0.47) Reference 0.59 (0.41, 0.84) 0.52 (0.36, 0.73) 0.51 (0.35, 0.75) 0.58 (0.31, 1.1) Sex Male Female Reference 0.71 (0.55, 0.90) Reference 0.54 (0.44, 0.67) Reference 0.47 (0.36, 0.60) Education < High school High school graduate Some college College graduate 7.8 (5.1, 12.0) 2.7 (1.9, 3.9) 1.8 (1.2, 2.7) Reference 2.5 (1.7, 3.6) 1.3 (0.95, 1.7) 1.2 (0.87, 1.6) Reference 2.1 (1.4, 3.1) 1.0 (0.75, 1.5) 1.2 (0.85, 1.7) Reference County New Castle Kent Sussex Reference 1.2 (0.88, 1.7) 1.7 (1.3, 2.3) Reference 1.2 (1.0, 1.7) 0.97 (0.73, 1.3) Reference 2.0 (1.4, 2.7) 1.3 (0.90, 1.8) Source: BRFSS, 2002. * In each column of table 12, the association between race/ethnicity and each marker of access to health care is examined after adjustment for age, sex, education, and county of residence. Therefore in the column for no health insurance, the results are interpreted as African Americans being 1.5 times more likely than Whites to lack health insurance. * The odds ratios measuring the effect of race were lower after adjusting for other predictors of access to health care, suggesting that some of the disparity observed in table 11 can be explained by the fact that minorities are more likely to have other factors that predict barriers to access to health care. However, since the race/ethnicity effect was not eliminated after adjustment for other variables, this indicates that the other variables in the model cannot account for all of the disparity observed. * The strongest predictor of not having each of the measures of access to health care was having less than a high school education. In tables 13 and 14, separate models were developed to examine this relationship with the following measures: 1) physical inactivity, 2) current smoking, 3) chronic drinking, 4) BMI greater than 25, and 5) eating fewer than five fruits and vegetables per day. Table 13. Association Between Race/Ethnicity and Modifiable Risk Factors for Cancer in Delaware Race/Ethnicity No Exercise Ever Smoked Obese (BMI: 25+) Poor Diet (Fewer Than Five Fruits and Vegetables/Day) Chronic Alcohol Intake1 White African American Hispanic Reference 2.2 (1.8, 2.7) 1.4 (0.94, 2.1) Reference 0.56 (0.46, 0.68) 0.65 (0.45, 0.94) Reference 2.2 (1.7, 2.7) 0.98 (0.67, 1.4) Reference 1.6 (1.2, 2.0) 0.90 (0.58, 1.4) Reference 0.64 (0.40, 1.0) 0.82 (0.35, 1.9) 1 Women who drank two or more drinks per day or men who drank three or more drinks per day. Source: BRFSS, 2002. * African Americans were more likely to report having no exercise, being obese, and eating fewer than five fruits and vegetables per day, and they were less likely to have ever smoked. * There were no significant differences observed between Hispanics and Whites. Table 14. Multivariate Regression Model of Predictors of Having Modifiable Risk Factors for Cancer in Delaware No Exercise Ever Smoked Obese (BMI: 25+) Poor Diet (Fewer Than Five Fruits and Vegetables/Day) Chronic Alcohol Intake1 Race/ethnicity White African American Hispanic Reference 2.4 (1.9, 3.0) 1.8 (1.1, 3.0) Reference 0.55 (0.45, 0.67) 0.70 (0.46, 1.1) Reference 2.5 (2.0, 3.2) 1.2 (0.75, 1.9) Reference 1.4 (1.1, 1.8) 0.71 (0.43, 1.2) Reference 0.54 (0.33, 0.88) 0.25 (0.06, 1.0) Age < 40 40–49 50–64 65–79 80+ Reference 1.3 (0.98, 1.6) 1.7 (1.3, 2.1) 2.4 (1.9, 3.1) 3.4 (2.4, 4.8) Reference 1.3 (1.1, 1.6) 1.9 (1.6, 2.2) 1.8 (1.5, 2.2) 0.82 (0.59, 1.1) Reference 1.4 (1.2, 1.7) 2.0 (1.6, 2.4) 1.7 (1.4, 2.1) 0.80 (0.58, 1.1) Reference 0.96 (0.77, 1.2) 0.92 (0.73, 1.2) 0.62 (0.49, 0.79) 0.57 (0.40, 0.82) Reference 0.88 (0.62, 1.3) 0.56 (0.38, 0.83) 0.52 (0.33, 0.81) 0.26 (0.09, 0.73) Sex Male Female Reference 1.8 (1.5, 2.2) Reference 0.65 (0.57, 0.74) Reference 0.47 (0.41, 0.54) Reference 0.48 (0.40, 0.57) Reference 0.66 (0.50, 0.87) Education < High school High school graduate Some college College graduate 4.4 (3.3, 5.9) 2.2 (1.8, 2.7) 1.6 (1.2, 2.0) Reference 2.2 (1.7, 2.8) 1.7 (1.4, 2.0) 1.5 (1.3, 1.8) Reference 1.9 (1.5, 2.6) 1.6 (1.3, 1.9) 1.4 (1.2, 1.7) Reference 2.3 (1.6, 3.2) 1.9 (1.6, 2.3) 1.6 (1.3, 1.9) Reference 1.1 (0.58, 2.0) 1.3 (0.89, 1.9) 1.4 (0.99, 2.1) Reference County New Castle Kent Sussex Reference 1.1 (0.88, 1.3) 0.95 (0.78, 1.2) Reference 1.0 (0.87, 1.2) 1.0 (0.85, 1.2) Reference 1.2 (1.0, 1.5) 1.2 (1.0, 1.4) Reference 1.1 (0.90, 1.3) 1.1 (0.90, 1.3) Reference 0.48 (0.34, 0.70) 0.74 (0.54, 1.0) Health insurance Yes No Reference 1.2 (0.85, 1.6) Reference 1.4 (1.1, 1.9) Reference 0.78 (0.58, 1.0) Reference 0.95 (0.68, 1.3) Reference 1.5 (0.95, 2.5) Personal doctor Yes No Reference 1.1 (0.80, 1.5) Reference 0.99 (0.76, 1.3) Reference 0.74 (0.56, 0.98) Reference 1.0 (0.76, 1.5) Reference 1.2 (0.73, 1.9) Usual source of care Yes No Reference 0.81 (0.57, 1.1) Reference 1.3 (0.96, 1.7) Reference 0.94 (0.71, 1.3) Reference 1.1 (0.80, 1.6) Reference 1.6 (0.98, 2.7) 1 Women who drank two or more drinks per day or men who drank three or more drinks per day. Source: BRFSS, 2002. * After adjusting for demographic characteristics and access to health care variables, the odds ratios for the associations between race/ethnicity and exercise and obesity were greater than the crude (i.e., unadjusted) odds ratios. This suggests that after including other potential predictors in the model, the effect between being African American and the risk of not exercising or being obese was stronger. * Having less than a high school education was a consistent strong predictor of being at risk for each of the five factors examined. Access to health care was not a predictor of being at risk for any factors, except not having health insurance increased the risk of ever smoking. Multivariate logistic regression models were developed to analyze whether the disparities in the use of cancer screening tests were due to race/ethnicity or were confounded by other factors, including socioeconomic status and access to health care. Separate models were developed for each screening test, and the outcome was the proportion of study respondents who did not receive the screening test within the recommended interval. The recommended time intervals for these analyses were: * A fecal occult blood test in the past year among individuals aged 50 or older * A sigmoidoscopy/colonoscopy in the past five years among individuals aged 50 or older * A prostate-specific antigen test in the past year for men aged 50 or older * A digital rectal exam in the past year for men aged 40 or older * A mammogram every two years for women aged 40 or older * A clinical breast exam in the past two years for women Odds ratios greater than one indicated that the subgroup was more likely not to have received the screening test within the recommended time interval than the reference group, and odds ratios less than one indicated that the subgroup was more likely to have received the screening test within the recommended interval. Table 15. Association Between Race and Not Receiving Cancer Screening Tests in Delaware Race Fecal Occult Blood Test Sigmoidoscopy or Colonoscopy Prostate- Specific Antigen Test Digital Rectal Exam Mammogram (Females) Clinical Breast Exam (Females) White African American Reference 1.4 (0.92, 2.1) Reference 1.3 (0.95, 1.8) Reference 0.68 (0.39, 1.2) Reference 0.78 (0.51, 1.2) Reference 0.83 (0.53, 1.3) Reference 0.78 (0.55, 1.1) Source: BRFSS, 2002. Table 16. Multivariate Model of Predictors for Not Receiving Cancer Screening Tests in Delaware Fecal Occult Blood Test Sigmoidoscopy or Colonoscopy Prostate- Specific Antigen Test Digital Rectal Exam Mammogram (Females) Clinical Breast Exam (Females) Race White African American Reference 1.2 (0.82, 1.9) Reference 1.1 (0.76, 1.5) Reference 0.49 (0.24, 1.0) Reference 0.62 (0.32, 1.0) Reference 0.93 (0.36, 1.1) Reference 0.70 (0.46, 1.1) Age < 40 40–49 50–64 65–79 80+ NA NA Reference 0.77 (0.60, 0.97) 1.0 (0.70, 1.6) NA NA Reference 0.68 (0.55, 0.84) 1.0 (0.73, 1.5) NA NA Reference 0.67 (0.46, 0.98) 0.89 (0.46, 1.7) NA Reference 0.29 (0.21, 0.40) 0.25 (0.17, 0.36) 0.38 (0.20, 0.73) NA Reference 0.35 (0.24, 0.51) 0.53 (0.36, 0.78) 1.4 (0.85, 2.2) Reference 1.5 (0.99, 2.2) 1.3 (0.90, 1.9) 2.1 (1.4, 3.0) 5.5 (3.4, 8.9) Sex Male Female Reference 1.0 (0.83, 1.3) Reference 1.1 (0.91, 1.4) NA NA NA NA NA NA NA NA Education < High school High school graduate Some college College graduate 1.2 (0.80, 1.8) 0.97 (0.73, 1.3) 0.85 (0.63, 1.2) Reference 2.7 (1.9, 3.9) 2.1 (1.6, 2.7) 1.6 (1.2, 2.1) Reference 1.8 (1.0, 3.2) 0.86 (0.55, 1.3) 0.99 (0.63, 1.6) Reference 2.3 (1.4, 3.7) 1.3 (0.94, 1.9) 0.94 (0.64, 1.4) Reference 2.1 (1.2, 3.6) 1.8 (1.2, 2.6) 1.2 (0.75, 1.8) Reference 4.5 (2.8, 7.4) 2.5 (1.7, 3.7) 2.0 (1.3, 3.0) Reference County New Castle Kent Sussex Reference 0.69 (0.51, 0.93) 0.60 (0.45, 0.78) Reference 1.0 (0.81, 1.4) 1.1 (0.88, 1.4) Reference 0.76 (0.48, 1.2) 0.92 (0.60, 1.4) Reference 0.76 (0.53, 1.1) 0.86 (0.62, 1.2) Reference 1.0 (0.68, 1.5) 1.2 (0.84, 1.7) Reference 0.73 (0.53, 1.0) 0.73 (0.54, 1.0) Health insurance Yes No Reference 1.4 (0.74, 2.7) Reference 2.5 (1.4, 4.6) Reference 3.7 (1.6, 8.3) Reference 2.2 (1.1, 4.3) Reference 2.6 (1.5, 4.6) Reference 2.1 (1.4, 3.2) Personal doctor Yes No Reference 2.5 (1.2, 5.3) Reference 2.3 (1.3, 4.0) Reference 3.8 (1.6, 8.6) Reference 4.0 (2.0, 8.1) Reference 3.5 (1.9, 6.4) Reference 1.7 (1.0, 2.8) Usual source of care Yes No Reference 0.72 (0.42, 1.2) Reference 0.93 (0.57, 1.5) Reference 0.85 (0.41, 1.8) Reference 1.1 (0.59, 1.9) Reference 1.9 (1.0, 3.5) Reference 2.2 (1.3, 3.5) Source: BRFSS, 2002. * A lack of health insurance or a personal doctor was associated with not receiving screening exams; lack of a usual source of care appeared to impact only receipt of a clinical breast exam. * Having less than a high school education was strongly associated with not receiving screening exams. * With the exception of clinical breast exams, advancing age was most often associated with an increasing likelihood of receipt of screening exams. * There was no association between race, sex or county of residence and the receipt of screening exams. 3.3.2. Stage at Diagnosis Stage of disease at site-specific cancer diagnosis was evaluated for the different racial and ethnic groups in Delaware by examining the proportion of individuals who were diagnosed at each cancer stage. The variations in stage at diagnosis for female breast, colorectal, and prostate cancer were examined by county and age at diagnosis. Stage was classified as local, regional, or distant. Regional and distant were combined to form the category of advanced stage. 3.3.2.1. Female Breast Cancer Figure 25. Percentage of female breast cancer cases, by stage at diagnosis and race, in Delaware Source: DCR, 1998–2002. * A larger proportion of White women were diagnosed with local disease than African American women, while more African American women were diagnosed with regional disease. White and African American women were equally likely to be diagnosed with distant disease. Figure 26. Percentage of female breast cancer cases diagnosed at an advanced stage, by race and county of residence, in Delaware Source: DCR, 1998–2002. * African American women residing in all counties in Delaware had a higher percentage of advanced disease than White women, with no strong evidence of differences depending on the county of residence at diagnosis. Figure 27. Percentage of female breast cancer cases diagnosed at an advanced stage, by race and age at diagnosis, in Delaware Source: DCR, 1998–2002. * African American women diagnosed with breast cancer when they were aged 79 or younger were more likely to be diagnosed with advanced disease than White women. * The effect of age at diagnosis on the proportion diagnosed with advanced disease may depend on race. For African American women the proportion with advanced disease generally decreased with age, while in White women the proportion decreased until age 70, when it began to increase. 3.3.2.2. Colorectal Cancer Figure 28. Percentage of colorectal cancer cases, by stage at diagnosis and race, in Delaware Source: DCR, 1998–2002. * African Americans were equally likely to be diagnosed with local disease as Whites. The data suggest that Whites may be slightly more likely to be diagnosed with regional disease, and African Americans with distant disease. Figure 29. Percentage of colorectal cancer cases diagnosed at an advanced stage, by race and county of residence, in Delaware Source: DCR, 1998–2002. * There is no evidence that the proportion of colorectal cancer cases diagnosed at advanced stage differs between Whites and African Americans. * The data suggest that for all races the proportion diagnosed with advanced disease was highest in New Castle County and lowest in Sussex County. Figure 30. Percentage of colorectal cancer cases diagnosed at an advanced stage, by race and age at diagnosis, in Delaware Source: DCR, 1998–2002. * The proportions of Whites diagnosed with advanced stage disease may decrease with age until after age 80, when it increases. * The proportion of African Americans diagnosed with advanced stage disease was not affected by age at diagnosis. 3.3.2.3. Prostate Cancer Figure 31. Percentage of prostate cancer cases, by stage at diagnosis and race, in Delaware Source: DCR, 1998–2002. * There was minimal difference between Whites and African Americans regarding the stage of disease at diagnosis. Figure 32. Percentage of prostate cancer cases diagnosed at an advanced stage, by race and county of residence, in Delaware Source: DCR, 1998–2002. * There was minimal difference in the proportion diagnosed with advanced stage disease, and it does not appear to differ by county of residence. Figure 33. Percentage of prostate cancer cases diagnosed at an advanced stage, by race and age at diagnosis, in Delaware Source: DCR, 1998–2002. * Men of both races aged 80 or older at diagnosis were the most likely to be diagnosed with advanced stage disease. * There was little difference in the proportion between Whites and African Americans. 3.3.3. Cancer Treatment Table 17 displays the results of the validation study undertaken to assess the completeness/ accuracy of the treatment-related data resident in the Delaware Cancer Registry; it reports the number and percentage of cancer cases receiving “standard” treatment (pre-validation study) and “appropriate” treatment (post-validation study). Table 17. Number and Percentage of Delaware Breast or Colorectal Cancer Patients With Local Disease Treated for Cancer in Delaware, by Receipt of Standard / Appropriate Cancer Treatment Study Female Breast # (%) Colorecta l # (%) Pre-validation study Received standard treatment Did not receive standard treatment 814 (75.5) 264 (24.5) 3 (4.1) 70 (95.9) Post-validation study Received appropriate treatment Did not receive appropriate treatment 1,001 (93.1) 74 ( 6.9) 63 (91.3) 6 (8.7) Source: DCR, 1998–2002 (Pre-validation); DCR + source documents (Post-validation) * The proportion of patients receiving standard/appropriate treatment changed dramatically between the pre- and post-validation studies: Receipt of appropriate breast cancer treatment increased from 75.5% of patients to 93.1%; receipt of appropriate colorectal cancer treatment increased from 4.1% to 91.3%. Source document review revealed substantial differences between the treatment documented in patients’ medical records and the treatment captured into the DCR. In the case of breast cancer treatment, these differences chiefly took the form of incomplete DCR data; e.g., radiation treatments received post-breast conserving surgery were clearly documented in source records but were not recorded in DCR files. In the case of colorectal cancer, the very substantial change in the proportion of patients receiving appropriate treatment was chiefly due to reviewers’ ability to determine that the surgery performed (i.e., polypectomy v. at least partial colectomy) was appropriate, given the specific site and stage of cancer reflected in the surgery and pathology reports. Table 18 presents the results of an analysis examining the likelihood of an association between receipt of standard treatment and each of three other variables: treating facility, type of cancer, and race. The analysis was done on both the pre-validation dataset and the post-validation dataset. Table 18. Association Between Receipt of Standard Treatment and Treating Facility, Cancer Type and Race Among Individuals with Local Cancer and Treated in Delaware Likelihood of Receipt of Standard Treatment Associated with Facility Associated with Cancer Type Associated with Race Pre-validation study Statistically significant p <0.0001 Statistically significant p <0.0001 Statistically significant p <0.0007 Post-validation study Statistically significant p <0.0001 Not statistically significant p = 0.57 Not statistically significant p = 0.33 Source: DCR, 1998–2002 (Pre-validation); DCR + source documents (Post-validation) * Treating facility, cancer type and race were all strongly associated with the likelihood of receiving standard treatment in the pre-validation study, which relied solely on DCR data. * In the post-validation analysis, which included source data not captured in the DCR, only treating facility continued to show a strong association with the likelihood of receiving appropriate treatment. 3.4. HEALTH POLICY, HEALTH SYSTEM, AND SOCIETAL FACTORS The impact of determinants of disparities that operate on an individual level has been examined in earlier sections of this report. However, research has suggested that some of the explanation for disparities may be the result of societal not individual factors. As we have seen from previous analyses, socioeconomic status is a significant factor affecting racial/ethnic disparities in the cancer burden in the United States. Unfortunately, many of the other societal components were not available in the data sources used. In this section the literature on these societal factors is briefly reviewed. This review considers two aspects of societal factors: patient and system barriers to preventive and cancer care. Patient barriers affect the benefit that minorities receive from the delivery of cancer services, whereas system barriers include factors that determine inequality in the delivery of the services. The relationship between race and ethnicity may be even more complex, as outlined in the model used by the National Center on Minority Health and Health Disparities in its Strategic Research Plan and Budget to Reduce and Ultimately Eliminate Health Disparities. Several patient barriers affect morbidity and mortality rates among different racial/ethnic populations. Financial and socioeconomic barriers, including poverty, lack of health insurance, and underinsurance, greatly influence access to appropriate early detection, treatment, and palliative care. Several studies have found that African Americans, especially those without insurance and with a lower household income, are more likely to be diagnosed with later stage cancer, which contributes to an increased mortality rate. Hiatt et al. found that the strongest predictors of cancer screening were private health insurance and frequent use of medical services (33). African Americans have a two-fold increased risk of being diagnosed with advanced-stage prostate cancer compared with non-Hispanic Whites (34). Roetzheim et al. (2000) examined the impact of insurance status among colorectal cancer patients in Florida and found that among non-Medicare patients, those with no health insurance were less likely to undergo surgery than patients with indemnity insurance, after adjustment for stage at diagnosis. Uninsured patients and patients enrolled in health maintenance organization (HMO) plans also had higher mortality rates than patients enrolled in fee-for-service plans. Physical barriers can also impede a patient’s ability to access proper care. Many patients lack transportation, are physically disabled, or are too frail to travel to a medical facility. Competing life demands can become a physical barrier when patients are unable to take time off from a job to receive needed services, particularly when providers lack convenient service hours. Other patient barriers result from a patient’s lack of health literacy or inability to assess information regarding cancer prevention and treatment. In minority communities there is often a lack of knowledge regarding the risks associated with not participating in recommended preventative screenings and not maintaining a healthy lifestyle, resulting in decreased screening rates in these communities. Even when minority communities are educated about cancer health, recommendations are often not acted on appropriately due to potential embarrassment and discomfort or to confusion about how to seek the needed care. Language barriers and a lack of translators prevent many minorities from accessing both information and treatment. The Breast and Cervical Cancer Intervention Study found low rates of Pap test screening, the key preventive measure against cervical cancer, among non-English-speaking Latina and Chinese women, a finding that may be generalized to apply to other immigrant populations (35). Language is not the only cultural barrier preventing minorities from receiving proper cancer care; a distrust of conventional medical care, fatalism, and modesty among minority communities also contribute to low screening rates (8). Like patient barriers, health system barriers affect cancer morbidity and mortality rates among minorities. Cancer services are not delivered as a lifetime continuum of care. Many minorities do not receive recommended cancer screenings, and those who do often receive inconsistent followup care. African Americans in some areas of the United States have higher Pap test rates but are still diagnosed in later stages of disease and have higher mortality than Whites. This can be attributed, in part, to the testing being performed in areas without an infrastructure for notification and followup of abnormal results (35). Followup care may be substandard among minority communities, since they may not have access to physicians with adequate training in adult or pediatric oncology or newer or proven therapies. End-of-life care tends to be worse among minority patients due to inequalities in pain management and hospice care; patients seen at outpatient centers that predominantly treat minorities were three times more likely than those treated elsewhere to have inadequate pain management (36). Studies have also shown lower usage of hospice care among minority communities, including African Americans, Asians, and White and non-White Hispanics (8). Further system barriers stem from failures among Federal, State, and local agencies in implementing or enforcing policies that address racial/ethnic disparities in cancer treatment and prevention. Title VI of the U.S. Civil Rights Act of 1964 prohibits discrimination and recipient practices that have the effect of discrimination on the basis of race, color, and national origin in programs and activities receiving Federal financial assistance (37). In addition to Title VI, national standards for culturally and linguistically appropriate services (CLAS) in health care were published in 2000 by the Department of Health and Human Services’ Office of Minority Health and adopted in an effort to implement standards of equality in health care for minorities (38). It is clear, however, that Title VI and the standards set in place in CLAS are not being upheld to the fullest extent possible, since minorities continue to be inadequately insured and face unequal access to quality cancer care. National efforts to combat these inequalities include the National Breast and Cervical Cancer Early Detection Program and the Screen for Life: National Colorectal Action Campaign, both administered by CDC. These programs are aimed at providing preventive education and screening for at-risk populations. Specifically, the National Breast and Cervical Cancer Early Detection Program helps low-income, uninsured, and underserved women gain access to lifesaving screening programs for early detection of female breast and cervical cancers, and the Screen for Life: National Colorectal Action Campaign informs men and women aged 50 years or older about the importance of having regular colorectal cancer screening tests (39, 40). These programs promote national awareness of the importance of cancer prevention and early detection. More can be done at the State level to increase access to cancer care and reduce cancer risks. Delaware has made progress in passing legislation mandating insurance coverage for annual mammograms for asymptomatic women aged 40 and over, ensuring that annual Pap tests are covered in health insurance contracts, and requiring group and individual health insurers, HMOs, and health service corporations to provide insurance coverage for colorectal cancer screening in certain situations (41). There are two areas where improvement is possible: 1) insurance coverage is ensured for only select phases of clinical trials and 2) the Tobacco Excise Tax is still below the national average. Increasing coverage for clinical trials would give the uninsured access to cutting-edge medical treatments and inform researchers of the efficacy of the treatments in a study group that is more representative of the population of cancer patients. Raising the tobacco excise tax would encourage residents of Delaware to quit smoking due to increased costs. Even with the presence of antidiscrimination laws, racial/ethnic discrimination still affects cancer service. Research shows that minorities receive a lower quality of care when they are in the health care system, after controlling for social determinants and insurance status. A phone survey of 3,884 people conducted by the Kaiser Family Foundation in 2000 found that 36 percent of Hispanics and 35 percent of African Americans (compared with 15 percent of Whites) felt that they were treated unfairly in the health care system based on their race/ethnicity (42). Research shows a lack of initiative among health care delivery systems to take on the challenge of eliminating patient and system barriers due to a high initial cost and lack of incentives (11). Further research is necessary to determine the best strategies for the Federal, State, and private sectors to implement in order to jointly address the elimination of cancer disparities. REFERENCES (1) Delaware Division of Public Health. Delaware Annual Vital Statistics Report 2002. 2004. (2) Delaware Division of Public Health. Delaware Cancer Incidence and Mortality Rates. 2005. (3) American Cancer Society. Mid Atlantic Division: Cancer Facts and Figures 2004. 2005. (4) American Cancer Society. American Cancer Society Cancer Facts and Figures for African Americans 1998–1999. 1998. Atlanta, GA. (5) American Cancer Society. American Cancer Society Cancer Facts and Figures 2005. 2005. Atlanta, GA. (6) Wingo P, Ries L, Giovino G, et al. 1999. Annual report to the nation on the status of cancer, 1973–1996, with a special section on lung cancer and tobacco smoking. J. Natl. Cancer Inst., 91, 675–690. (7) Singh GK, Miller BA, Hankey BF, Edwards BK. Area socio-economic variations in US cancer incidence, mortality, stage, treatment, and survival 1975–1999. 4. 2003. Bethesda, MD, National Cancer Institute. NIH Publication No. 03-5417. NCI Cancer Surveillance Monograph Series. (8) Ward E, Jemal A, Cokkinides V, et al. 2004. Cancer disparities by race/ethnicity and socioeconomic status. CA Cancer J Clin, Mar–Apr:54(2), 78–93. (9) American Cancer Society. Cancer Facts and Figures for African Americans 2005–2006. 2005. (10) Institute of Medicine. Unequal Treatment: Confronting Racial and Ethnic Disparities in Healthcare. 34. 2002. Washington, DC, National Academy Press. (11) Institute of Medicine. The Unequal Burden of Cancer: An Assessment of NIH Research and Programs for Ethnic Minorities and the Medically Underserved. 1999. Washington, DC, National Academy Press. (12) Epstein A, Ayanian J. 2001. Racial disparities in medical care. N Engl J Med, 344, 1471– 1472. (13) National Institutes of Health. Strategic Research Plan and Budget to Reduce and Ultimately Eliminate Health Disparities. 2002. (14) Division of Adult and Community Health, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, Behavioral Risk Factor Surveillance System Online Prevalence Data, 1995–2004. www.cdc.gov/brfss. 2005. (15) Centers for Disease Control and Prevention. Health, United States 1998. Lung Cancer Mortality. 1998. (16) Centers for Disease Control and Prevention. Health, United States 1998. Population, Education. 1998. (17) Centers for Disease Control and Prevention. Health Status, Causes of Death. 2005. (18) DeVoe JE, Fryer GE, Phillips R, Green L. 2003. Receipt of preventive care among adults: Insurance status and usual source of care. Am J Public Health, 93(5), 786–791. (19) SEER*Stat. National Cancer Institute, 2005. (20) 62 Del. Laws c.334 §1; 70 Del. Laws c391 §1. Delaware Cancer Control Act. 1980. (21) National Cancer Institute. SEER Registries. http://www.seer.cancer.gov/registries/. 2005. (22) U.S. Census Bureau. www.census.gov. 2005. (23) U.S. Population Data. http://www.seer.cancer.gov/popdata/download.html. December 13, 2004. (24) Delaware Population Consortium. http://www.cadsr.udel.edu/demography/consortium.htm. 2005. (25) SEER ICD-O-3 Coding Materials. http://www.seer.cancer.gov/icd-o-3/. 2005. (26) National Program of Cancer Registries. Submission Guidelines. https://www.npcrcss.org/ utilities/. 2005. (27) Kleibaum DG, Kupper LL, Morgenstern H. Epidemiologic Research: Principles and Quantitative Methods. Van Norstrand Reinhold, 1982. (28) EpiBasic. http://www.biostat.au.dk/teaching/software/. 2005. (29) SAS. Cary, NC. 2003. (30) U.S. Department of Health and Human Services. Healthy People 2010: Understanding and Improving Health. 2nd ed. Washington, D.C. 2000. (31) North American Association of Central Cancer Registries. Standards for Cancer Registries Volume II: Data Standards and Data Dictionary Version 10.1. 8th ed. Springfield, IL. 2003. (32) Shavers VL, Brown ML. 2002. Racial and ethnic disparities in the receipt of cancer treatment. J Natl Cancer Inst, 94(5), 334–357. (33) Hiatt RA, Pasick RJ, Stewart S, et al. 2001. Community-based cancer screening for undeserved women: Design and baseline findings from the Breast and Cervical Cancer Intervention Study. Preventive Medicine, 33, 190–203. (34) Hoffman RM, Gilliland FD, Eley JW, et al. 2001. Racial and ethnic differences in advanced-stage prostate cancer: the Prostate Cancer Outcomes Study. J Natl Cancer Inst, 93(5), 388–395. (35) Garner E. 2003. Cervical cancer: Disparities in screening, treatment, and survival. Cancer Epidemiology Biomarkers and Prevention, 12, 242s–247s. (36) Cleeland CS, Gonin R, Hatfield AK, et al. 1994. Pain and its treatment in outpatients with metastatic cancer. N Engl J Med., 330(9), 592–596. (37) United States Supreme Court. Title VI of the Civil Rights Act of 1964. 42, 200d et seq. 1964. (38) Office of Minority Health. National Standards for Culturally and Linguistically Appropriate Services in Health Care. U.S. Department of Health and Human Services, editor. 2001. (39) National Center for Chronic Disease and Prevention. The National Breast and Cervical Cancer Early Detection Program: Saving Lives Through Screening. Centers for Disease Control and Prevention, editor. www.cdc.gov. 2005. (40) National Center for Chronic Disease and Prevention. Screen For Life: National Colorectal Cancer Action Campaign. www.cdc.gov. 2005. Centers for Disease Control and Prevention. (41) National Center for Chronic Disease and Prevention. Cancer Legislative Information. www.cdc.gov. 2005. Centers for Disease Control and Prevention. (42) Kaiser Family Foundation. Race, Ethnicity, and Medical Care. Kaiser Family Foundation. 2005. (43) Singh GK, Miller BA, Hankey BF, Edwards BK. Area Socioeconomic Variations in U.S. Cancer Incidence, Mortality, Stage, Treatment, and Survival. 1975–1999. 4. 2003. Bethesda, MD, National Cancer Institute. National Cancer Surveillance Monograph Series. (44) Bach PB, Cramer LD, Warren JL, Begg CB. 1999. Racial differences in the treatment of early-stage lung cancer. N Engl J Med, 341(16), 1198–1205. (45) Merrill RM, Merrill AV, Mayer LS. 2000. Factors associated with no surgery or radiation therapy for invasive cervical cancer in Black and White women. Ethn Dis, 10(2), 248–256. APPENDIX A: SUPPLEMENTAL INFORMATION ON REPORT METHODOLOGY Table A1. NAACCR Certification of the Delaware Cancer Registry’s 2002 Incidence Data NAACCR Registry Certification on Quality, Completeness, and Timeliness of 2002 Data Summary of Certification Measures Registry Element Gold Standard Silver Standard Actual Measure Measurement Error Allowed Standard Achieved 1. Completeness of case ascertainment 95% 90% 92.2% 1.0% Silver 2. Completeness of information recorded * Missing/unknown “age at diagnosis” * Missing/unknown “sex” * Missing/unknown “race” Missing/unknown “state/province & county” <=2% <=2% <=3% <=2% <=3% <=3% <=5% <=3% 0.0% 0.1% 1.2% 0.1% -0.4% -0.4% -0.4% -0.4% Gold Gold Gold Gold 3. Death certificate only cases <=3% <=5% 1.2% -0.4% Gold 4. Duplicate primary cases <=1 per 1,000 <=2 per 1,000 0.0 per 1,000 -0.4 per 1,000 Gold 5. Passing edits 100% 97% 100% Not applicable Gold 6. Timeliness Data submitted within 24 months of close of accession year Gold Certification Status Silver Table A2. Number and Percentage of Cancer Cases, by Race/Ethnicity and Cancer Site, Delaware 1998–2002 Race/Ethnicity All Female Breast Colorectal Lung and Bronchus Prostate Total White African American Hispanic Asian, Pacific Islanders, American Indians, and Alaska Natives Missing 19,849 16,416 (83) 2,867 (14) 187 (0.9) 169 (0.8) 210 (1.1) 2,792 2,322 (83) 392 (14) 23 (0.8) 33 (1.2) 22 (0.8) 2,212 1,841 (83) 324 (15) 9 (0.4) 20 (0.9) 18 (0.8) 3,070 2,609 (85) 418 (14) 18 (0.6) 19 (0.6) 6 (0.2) 3,134 2,448 (78) 591 (19) 24 (0.8) 16 (0.5) 55 (1.8) Table A3. Comparison of the Percentage of Individuals in Each Stage of Diagnosis and Race Group in Delaware and the United States, 1998–2002 Delaware United States White African American White African American Female breast cancer Local Regional Distant Unstaged 67 26 4 4 58 33 4 5 65 29 4 2 53 36 7 4 Colorectal cancer Local Regional Distant Unstaged 29 47 16 7 29 44 21 6 40 38 17 5 35 35 23 7 Prostate cancer Local Regional Distant Unstaged 83 8 4 5 85 5 4 6 80 13 4 3 79 12 6 3 Table A4. Description of the SEER Program Codes and Surgical Procedures Used To Define Standard Treatment for Each of the Cancers Examined Female Breast Colorectal Lung and Bronchus Prostate 0 No surgery 10-19 Local tumor destruction 20-29 Partial mastectomy Local tumor excision Excision of <1 lobe Local tumor excision 30-39 Subcutaneous mastectomy Partial colectomy Partial pneumonectomy Subtotal prostatectomy 40-49 Total mastectomy Hemi-colectomy Extended lobe or bilobectomy 50-59 Modified radical mastectomy Total colectomy Pneumonectomy Total prostatectomy 60-69 Radical mastectomy Total proctocolectomy Extended pneumonectomy 70-79 Extended radical mastectomy 30-60 with contiguous organs Extended radical pneumonenctomy 50 with contiguous organs 80 Mastectomy, NOS Colectomy, NOS Resection, NOS Prostatectomy, NOS 90 Surgery, NOS 99 Unknown Cells in blue designate standard treatment, and cells in yellow refer to nonstandard treatment. Cells in green were defined as standard if individuals also received radiation; if they did not receive radiation, they were coded as not receiving standard treatment. The cells in gray were considered to be missing data on surgical procedure. APPENDIX B: SUPPLEMENTAL INCIDENCE AND MORTALITY DATA Table B1. Number of New Cancer Cases and Age-Adjusted Incidence Rates (With 95-Percent Confidence Intervals), by Race in Delaware and the United States, 1998–2002 White African American Number Rate (95% CI) Number Rate (95% CI) Delaware All sites Female breast Colorectal Lung and bronchus Prostate 16,416 2,322 1,841 2,609 2,448 489.9 (482.5, 497.5) 130.2 (125.0, 135.6) 54.5 (52.1. 57.1) 76.1 (73.2, 79.1) 159.4 (153.2, 165.9) 2,867 392 324 418 591 531.4 (512.3, 551.3) 117.6 (106.5, 129.9) 64.9 (58.2, 72.4) 80.7 (73.3, 88.8) 267.6 (246.9, 290.1) United States All sites Female breast Colorectal Lung and bronchus Prostate 508,532 80,260 55,976 67,118 78,732 489.1 (487.7, 490.4) 143.0 (142.0, 144.0) 53.6 (53.1, 54.0) 65.0 (64.5, 65.4) 173.5 (172.3, 174.7) 55,966 6,362 8,421 11,639 7,907 519.3 (514.9, 523.8) 118.7 (116.1, 121.4) 62.6 (61.1, 64.2) 81.7 (79.9, 83.5) 277.1 (271.9, 282.5) Rates are per 100,000 and age-adjusted to the 2000 U.S. standard million population; rates based on counts of fewer than 25 are suppressed. Sources: Delaware Department of Health, Delaware Cancer Registry, 2002; SEER Program, based on the November 2003 submission; and National Center for Health Statistics. There were 150 cases of cancer diagnosed among Asians and Pacific Islanders, with a rate of 282.6 (240.8, 331.7), and 187 cases of cancer diagnosed among Hispanics, with a rate of 250.4 (217.0, 289.0). Table B2. Number of Cancer Deaths and Age-Adjusted Mortality Rates (With 95-Percent Confidence Intervals), by Race in Delaware and the United States, 1998–2002 White African American Number Rate (95% CI) Number Rate (95% CI) Delaware All sites Female breast Colorectal Lung and bronchus Prostate 7,002 513 676 2,128 328 207.5 (202.6, 212.4) 27.2 (25.0, 29.7) 20.0 (18.6, 21.6) 62.1 (59.5, 64.8) 26.3 (23.5, 29.2) 1,249 115 140 335 102 251.3 (237.8, 265.6) 36.2 (30.1, 43.4) 29.4 (24.9, 34.7) 66.8 (60.0, 74.3) 64.9 (53.5, 78.8) United States All sites Female breast Colorectal Lung and bronchus Prostate 2,389,534 176,758 245,335 682,251 127,744 195.3 (195.0, 195.5) 25.9 (25.8, 26.0) 20.0 (19.9, 20.0) 55.8 (55.6, 55.9) 27.7 (27.6, 27.9) 309,865 27,078 33,963 80,407 26,549 248.1 (247.3, 249.0) 34.7 (34.3, 35.1) 27.9 (27.6, 28.2) 64.1 (63.6, 64.5) 68.1 (67.3, 69.0) Rates are per 100,000 and age-adjusted to the 2000 U.S. standard million population; rates based on counts of fewer than 25 are suppressed. Sources: Delaware Department of Health, Delaware Cancer Registry, 2002; SEER Program, based on the November 2003 submission; and National Center for Health Statistics. In Delaware, there were 42 deaths from cancer among Asians and Pacific Islanders, with a rate of 104.7 (77.4, 141.6), and there were 80 deaths from cancer among Hispanics, with a rate of 149.9 (120.4, 186.6). Table B3. Number of New Cancer Cases and Age-Adjusted Incidence Rates (With 95-Percent Confidence Intervals), by Race and Sex in Delaware and the United States, 1998–2002 White African American Number Rate (95% CI) Number Rate (95% CI) Delaware All sites Male Female Colorectal Male Female Lung and bronchus Male Female 8,524 7,892 932 909 1,467 1,142 570.3 (558.3, 582.5) 433.9 (424.4, 443.6) 63.9 (59.9, 68.2) 47.0 (44.1, 50.2) 97.1 (92.3, 102.2) 60.3 (56.9, 63.8) 1,522 1,343 149 174 234 194 679.3 (646.0, 714.3) 426.3 (404.1, 449.7) 71.3 (60.7, 83.7) 59.6 (51.3, 69.1) 108.2 (95.2, 123.0) 65.6 (57.0, 75.5) United States All sites Male Female Colorectal Male Female Lung and bronchus Male Female 259,777 248,755 28,068 27,908 36,430 30,688 569.7 (567.5, 571.9) 435.6 (433.9, 437.3) 62.9 (62.2, 63.6) 46.4 (45.8, 46.9) 81.3 (80.4, 82.1) 53.4 (52.8, 54.0) 30,216 25,750 2,995 3,367 4,957 3,464 692.5 (684.3, 700.9) 404.1 (399.1, 409.2) 72.9 (70.2, 75.8) 56.2 (54.3, 58.1) 117.4 (114.0, 121.0) 57.6 (55.7, 59.6) Rates are per 100,000 and age-adjusted to the 2000 U.S. standard million population; rates based on counts of fewer than 25 are suppressed. Sources: Delaware Department of Health, Delaware Cancer Registry, 2002; SEER Program, based on the November 2003 submission; and National Center for Health Statistics. Table B4. Number of Cancer Deaths and Age-Adjusted Mortality Rates (With 95-Percent Confidence Intervals), by Race and Sex in Delaware and the United States, 1998–2002 White African American Number Rate (95% CI) Number Rate (95% CI) Delaware All sites Male Female Colorectal Male Female Lung and bronchus Male Female 3,615 3,387 349 327 1,220 908 253.9 (245.8, 262.3) 176.4 (170.6, 182.5) 24.8 (22.3, 27.5) 16.6 (14.9, 18.5) 82.5 (78.0, 87.2) 47.2 (44.2, 50.4) 634 615 64 76 198 137 329.8 (305.1, 356.5) 206.2 (190.5, 223.2) 34.0 (26.6, 43.4) 26.3 (21.0, 32.9) 97.4 (84.7, 111.9) 46.8 (39.6, 55.4) United States All sites Male Female Colorectal Male Female Lung and bronchus Male Female 1,236,918 1,152,616 122,299 123,036 393,149 289,102 242.5 (242.0, 242.9) 164.5 (164.2, 164.8) 24.3 (24.1, 24.4) 16.8 (16.7, 16.9) 75.2 (75.0, 75.5) 41.8 (41.7, 42.0) 163,486 146,379 16,175 17,788 50,738 29,669 339.4 (337.6, 341.1) 194.3 (193.3, 195.3) 34.0 (33.4, 34.5) 24.1 (23.7, 24.4) 101.3 (100.3, 102.2) 39.9 (39.4, 40.3) Rates are per 100,000 and age-adjusted to the 2000 U.S. standard million population; rates based on counts of fewer than 25 are suppressed. Sources: Delaware Department of Health, Delaware Cancer Registry, 2002; SEER Program, based on the November 2003 submission; and National Center for Health Statistics. There were 41 deaths from cancer in Hispanic men, with a rate of 196.7, and 39 deaths from cancer among Hispanic women, with a rate of 121.9. Table B5. Number of New Cancer Cases and Age-Adjusted Incidence Rates (With 95-Percent Confidence Intervals), by Race and Age in Delaware, 1998–2002 White African American Age (Years) Number Rate (95% CI) Number Rate (95% CI) All sites 20–39 40–49 50–64 65–79 80+ 826 1,257 4,532 7,307 2,494 55.1 (51.5, 59.0) 280.1 (265.0, 296.0) 922.5 (896.1, 949.8) 2,185.3 (2135.8, 2236.0) 2,308.1 (2219.3, 2400.5) 238 361 984 980 304 47.6 (41.9, 54.1) 327.2 (295.2, 362.8) 1,065.1 (1000.6, 1133.8) 2,216.1 (2081.6, 2359.3) 2,450.0 (2189.5, 2741.5) Female breast 20–39 40–49 50–64 65–79 80+ 99 340 772 824 287 13.3 (10.9, 16.2) 149.8 (134.7, 166.6) 303.2 (282.5, 325.3) 449.0 (419.3, 480.7) 401.4 (357.5, 450.6) 46 112 122 80 32 18.0 (13.5, 24.1) 185.3 (154.0, 223.0) 244.8 (205.0, 292.3) 312.6 (251.1, 389.2) 375.8 (265.7, 531.4) Colorectal 20–39 40–49 50–64 65–79 80+ 30 94 391 875 451 2.0 (1.4, 2.9) 20.9 (17.1, 25.6) 79.6 (72.1, 87.9) 261.7 (244.9, 279.6) 417.4 (380.6, 457.7) < 25 33 87 123 66 NA 29.9 (21.3, 42.1) 94.2 (76.3, 116.2) 278.1 (233.1, 331.9) 531.9 (417.9, 677.0) Lung and bronchus 20–39 40–49 50–64 65–79 80+ 36 109 714 1,399 351 2.4 (1.7, 3.3) 24.3 (20.1, 29.3) 145.3 (135.1, 156.4) 418.4 (397.0, 440.9) 324.8 (292.6, 360.7) < 25 40 161 173 39 NA 36.3 (26.6, 49.4) 174.3 (149.3, 203.4) 391.2 (337.0, 454.1) 314.3 (229.6, 430.2) Prostate 20–39 40–49 50–64 65–79 80+ < 25 36 775 1,365 271 NA 16.2 (11.7, 22.5) 327.5 (305.3, 351.4) 904.9 (858.2, 954.2) 741.5 (658.3, 835.2) < 25 < 25 270 253 41 NA NA 634.6 (563.2, 715.0) 1,358.0 (1200.6, 1536.1) 1,053.4 (775.7, 1430.7) NA = Rates are based on counts too small to be displayed. Rates are per 100,000 and age-adjusted to the 2000 U.S. standard million population; rates based on counts of fewer than 25 are suppressed. Sources: Delaware Department of Health, Delaware Cancer Registry, 2002; SEER Program, based on the November 2003 submission; and National Center for Health Statistics. Table B6. Number of Cancer Deaths and Age-Adjusted Mortality Rates (With 95-Percent Confidence Intervals), by Race and Age in Delaware, 1998–2002 White African American Age (Years) Number Rate (95% CI) Number Rate (95% CI) All sites 20–39 40–49 50–64 65–79 80+ 142 303 1,450 3,302 1,805 9.5 (8.0, 11.2) 67.5 (60.3, 75.6) 295.2 (280.4, 310.8) 987.5 (954.4, 1021.8) 1,670.5 (1,595.2, 1,749.4) 47 114 340 515 233 9.4 (7.1, 12.5) 103.3 (86.0, 124.2) 368.0 (330.9, 409.3) 1,164.6 (1,068.2, 1,269.6) 1,877.8 (1,651.5, 2,135.1) Female breast 20–39 40–49 50–64 65–79 80+ < 25 39 118 196 147 NA 17.2 (12.6, 23.5) 46.3 (38.7, 55.5) 106.8 (92.8, 122.8) 205.6 (174.9, 241.6) < 25 25 41 29 < 25 NA 41.4 (28.0, 61.2) 82.3 (60.6, 111.7) 113.3 (78.8, 163.1) NA Colorectal 20–39 40–49 50–64 65–79 80+ < 25 < 25 121 323 206 NA NA 24.6 (20.6, 29.4) 96.6 (86.6, 107.7) 190.6 (166.3, 218.5) < 25 < 25 37 59 32 NA NA 40.1 (29.0, 55.3) 133.4 (103.4, 172.2) 257.9 (182.4, 364.7) Lung and bronchus 20–39 40–49 50–64 65–79 80+ < 25 67 519 1,142 384 NA 14.9 (11.8, 19.0) 105.6 (96.9, 115.1) 341.5 (322.3, 361.9) 355.4 (321.6, 392.8) < 25 25 112 152 44 NA 22.7 (15.3, 33.5) 121.2 (100.7, 145.9) 343.7 (293.2, 403.0) 354.6 (263.9, 476.5) Prostate 20–39 40–49 50–64 65–79 80+ < 25 < 25 < 25 150 157 NA NA NA 99.4 (84.7, 116.7) 429.6 (367.4, 502.3) <25 <25 <25 56 36 NA NA NA 300.6 (231.3, 390.6) 925.0 (667.2, 1,282.3) NA = Rates are based on counts too small to be displayed. Rates are per 100,000 and age-adjusted to the 2000 U.S. standard million population; rates based on counts of fewer than 25 are suppressed. Sources: Delaware Department of Health, Delaware Cancer Registry, 2002; SEER Program, based on the November 2003 submission; and National Center for Health Statistics. Table B7. Number of New Cancer Cases and Age-Adjusted Incidence Rates (With 95-Percent Confidence Intervals), by Race and County of Residence in Delaware, 1998–2002 White African American Number Rate (95% CI) Number Rate (95% CI) All sites Kent New Castle Sussex 2,286 9,661 4,461 467.4 (448.6, 487.0) 500.1 (490.2, 510.2) 481.3 (467.4, 495.6) 439 1,957 470 463.9 (422.5, 509.4) 557.2 (533.1, 582.5) 501.4 (458.0, 548.8) Female breast Kent New Castle Sussex 331 1,397 573 125.4 (112.6, 139.6) 133.2 (126.4, 140.4) 120.8 (111.3, 131.1) 55 284 53 98.8 (75.8, 128.7) 128.1 (114.1, 143.9) 96.6 (73.8, 126.4) Colorectal Kent New Castle Sussex 291 1,074 517 60.2 (53.6, 67.5) 54.6 (50.4, 58.6) 54.5 (50.0, 59.4) 68 210 46 78.1 (61.6, 99.1) 66.4 (58.0, 76.0) 51.0 (38.2, 68.1) Lung and bronchus Kent New Castle Sussex 422 1,394 793 85.4 (77.6, 93.9) 71.7 (68.0, 75.5) 80.5 (75.1, 86.3) 60 283 75 66.2 (51.4, 85.2) 83.8 (74.6, 94.1) 81.5 (65.0, 102.1) Prostate Kent New Castle Sussex 286 1,542 620 126.1 (112.3, 141.6) 182.5 (173.6, 191.9) 132.8 (122.7, 143.6) 91 411 88 210.4 (171.4, 258.4) 297.2 (269.8, 327.4) 227.0 (184.2, 279.8) Rates are per 100,000 and age-adjusted to the 2000 U.S. standard million population; rates based on counts of fewer than 25 are suppressed. Sources: Delaware Department of Health, Delaware Cancer Registry, 2002; SEER Program, based on the November 2003 submission; and National Center for Health Statistics. Table B8. Number of Cancer Deaths and Age-Adjusted Mortality Rates (With 95-Percent Confidence Intervals), by Race and County of Residence in Delaware, 1998–2002 White African American Number Rate (95% CI) Number Rate (95% CI) All sites Kent New Castle Sussex 1,048 4,055 1,899 215.7 (203.0, 229.2) 208.9 (202.6, 215.4) 202.4 (193.5, 211.8) 214 806 229 246.7 (215.8, 282.1) 252.1 (235.3, 270.1) 251.1 (220.6, 285.8) Female breast Kent New Castle Sussex 76 315 122 28.0 (22.3, 35.0) 28.2 (25.2, 31.5) 24.9 (20.8, 29.7) < 25 76 25 NA 36.8 (29.4, 46.1) 45.9 (31.0, 67.9) Colorectal Kent New Castle Sussex 81 401 194 16.8 (13.5, 20.9) 20.7 (18.7, 22.8) 20.7 (18.0, 23.8) 26 86 28 31.3 (21.3, 46.0) 28.9 (23.4, 35.7) 31.1 (21.5, 45.1) Lung and bronchus Kent New Castle Sussex 363 1,148 617 73.8 (66.6, 81.8) 58.9 (55.6, 62.4) 63.2 (58.4, 68.4) 51 219 65 57.0 (43.3, 75.0) 67.8 (59.4, 77.4) 71.0 (55.7, 90.5) Prostate Kent New Castle Sussex 37 199 92 20.9 (15.2, 28.9) 28.3 (24.6, 32.5) 24.4 (19.9, 30.0) < 25 59 < 25 NA 66.3 (41.6, 72.8) NA NA = Rates are based on counts too small to be displayed. Rates are per 100,000 and age-adjusted to the 2000 U.S. standard million population; rates based on counts of fewer than 25 are suppressed. Sources: Delaware Department of Health, Delaware Cancer Registry, 2002; SEER Program, based on the November 2003 submission; and National Center for Health Statistics. Table B9. Rate Ratios and Rate Difference Measuring Disparities in Five-Year Average Cancer Incidence Rates in Three Time Periods Between 1980 and 2002 1980–1984 1990–1994 1998–2002 Rate Ratio Rate Difference Rate Ratio Rate Difference Rate Ratio Rate Difference All sites Male Female 1.3 1.0 127.3 14.8 1.4 1.0 253.2 4.3 1.2 1.0 93.9 -8.2 Female breast 0.89 -11.6 0.90 -14.4 0.89 -14.1 Colorectal Male Female 0.7 0.8 -26.4 -11.9 1.0 1.1 0.9 6.8 1.1 1.3 8.4 12.7 Lung and bronchus Male Female 1.5 1.3 61.7 13.0 1.6 1.1 67.0 8.8 1.1 1.0 10.7 0.1 Prostate 2.0 73.5 1.7 137 1.6 100.7 Table B10. Rate Ratios and Rate Difference Measuring Disparities in Five-Year Average Cancer Mortality Rates in Three Time Periods Between 1980 and 2002 1980–1984 1990–1994 1998–2002 Rate Ratio Rate Difference Rate Ratio Rate Difference Rate Ratio Rate Difference All sites Male Female 1.5 1.3 139.4 47.7 1.6 1.3 184.7 56.6 1.3 1.2 75.9 30.3 Female breast 1.0 1.6 1.0 1.5 1.3 9.0 Colorectal Male Female 1.3 1.0 10.2 -0.8 1.1 1.4 2.6 8.9 1.4 1.5 8.8 9.1 Lung and bronchus Male Female 1.5 1.6 47.2 17.9 1.6 1.2 60.5 10.3 1.2 1.0 16.7 1.0 Prostate 2.2 40.2 2.5 56.2 2.4 37.7 APPENDIX C: SUPPLEMENTAL BRFSS INFORMATION Table C1. Wording of the BRFSS Questions for Each of the Variables Used in This Analysis Demographic Variables What county do you live in? Kent Sussex New Castle Don’t know/not sure Refused What is your age? Indicate sex of respondent. (Ask only if necessary) Male Female Are you Hispanic or Latino? Yes No Don’t know/not sure Refused Which one of the following groups would you say best represents your race? White Black or African American Asian Native Hawaiian or Other Pacific Islander American Indian, Alaska Native Other (specify) _____________________ Don’t know/not sure Refused What is the highest grade or year of school you completed? (Read as necessary) Never attended school or only attended kindergarten Grades 1 through 8 (elementary school) Grades 9 through 11 (some high school) Grade 12 or GED (high school graduate) College 1 year to 3 years (some college or technical school) College 4 years or more (college graduate) Refused Is your annual household income from all sources: 01 Less than $25,000 (If "no," skip to 5, if "yes," ask 2) 02 Less than $20,000 (If "no," code 1, if "yes," ask 3) 03 Less than $15,000 (If "no," code 2, if "yes," ask 4) 04 Less than $10,000 (If "no," code 2, if “yes” code 4) 05 Less than $35,000 (If "no," ask 6, if “yes” code 5) 06 Less than $50,000 (If "no," ask 7, if “yes” code 6) 07 Less than $75,000 (If "no," code 8, if “yes” code 7) 08 $75,000 or more 77 Don’t remember / Don’t know 99 Refused Access to Health Care Variables Do you have any kind of health care coverage, including health insurance, prepaid plans such as HMOs, or government plans such as Medicare? Yes No Don’t know/not sure Refused Do you have one person you think of as your personal doctor or health care provider? Yes, only one Yes, more than one No Don’t know/not sure Refused In the year prior to your cancer diagnosis, when you were sick or needed advice about your health, to which one of the following places did you usually go? A doctor’s office A public health clinic or community health center A hospital outpatient department A hospital emergency room Urgent care center Some other kind of place (specify) _______________ No usual place Don’t remember / Don’t know Refused Modifiable Risk Factor Variables When you are at work, which of the following best describes what you do? Mostly sitting or standing Mostly walking Mostly heavy labor or physically demanding work Don’t know/not sure Refused During the past month, other than your regular job, did you participate in any physical activities or exercise, such as running, calisthenics, golf, gardening, or walking for exercise? Yes No Don’t know/not sure Refused Have you smoked at least 100 cigarettes in your entire life? Yes No Don’t know/not sure Refused Do you now smoke cigarettes every day, some days, or not at all? Every day Some days Not at all Refused About how much do you weigh without shoes? __ __ __ Weight in pounds 7 7 7 Don’t know/not sure 9 9 9 Refused About how tall are you without shoes? __/__ __ Height in ft/inches 7 7 7 Don’t know/not sure 9 9 9 Refused The number of servings of fruit and vegetables was calculated by asking respondents each of the following questions individually using the response categories below. How often do you drink fruit juices such as orange, grapefruit, or tomato? Not counting juice, how often do you eat fruit? How often do you eat green salad? How often do you eat potatoes, not including French fries, fried potatoes, or potato chips? How often do you eat carrots? Not counting carrots, potatoes, or salad, how many servings of vegetables do you usually eat? 1 __ __ per day 2 __ __ per week 3 __ __ per month 4 __ __ per year 5 __ __ never 7 7 7 Don’t know/not sure 9 9 9 Refused A drink of alcohol is one 12 oz. can or bottle of beer, one glass of wine, one 12 oz. can or bottle of wine cooler, 1 cocktail, or 1 shot of liquor. During the past 30 days, how many days per week or per month did you have at least 1 drink of any alcoholic beverage? __ __ Days per week (specify number of days) __ __ Days in 30 prior to diagnosis (specify number of days) No drinks in past 30 days Don’t know/not sure Refused On the days when you drank, about how many drinks did you drink on the average? __ __ Number of drinks Don’t know/not sure Refused Screening Test Variables The questions for each of the six screening tests were worded similarly. First the respondent was asked if they had ever received the test then when they had last received the test. The explanation for each test and the response categories are listed below. A blood stool test is a test that may use a special kit at home to determine whether the stool contains blood. Have you ever had this test with a home kit? Sigmoidoscopy and colonoscopy are exams in which a tube is inserted in the rectum to view the bowel for signs of cancer or other health problems. Have you ever had either of these exams? A prostate-specific antigen test, also called a PSA test, is a blood test used to check men for prostate cancer. Have you ever had a PSA test? A digital rectal exam is an exam in which a doctor, nurse, or other health professional places a gloved finger into the rectum to feel the size, shape, and hardness of the prostate gland. Have you ever had a digital rectal exam? A clinical breast exam is when a doctor, nurse, or other health professional feels the breast for lumps. Have you ever had a clinical breast exam? A mammogram is an x-ray of each breast to look for breast cancer. Have you ever had a mammogram? Yes No Don’t know/not sure Refused How long has it been since you had your last blood stool test using a home kit? How long has it been since you had your last sigmoidoscopy or colonoscopy? How long has it been since you had your last PSA test? How long has it been since you had your last digital rectal exam? How long has it been since you had your last mammogram? How long had it been since your last clinical breast exam? Within the past year (any time less than 12 months ago) Within the past 2 years (1 year but less than 2 years ago) Within the past 5 years (2 years but less than 5 years ago) 5 or more years ago Don’t know/not sure Refused Table C2. Number of Individuals With Valid Data Who Were Included in the Denominator for Each Variable Analyzed, by Race/Ethnicity in Delaware and the United States Delaware United States White African American Hispanic White African American Hispanic Health insurance 3,228 508 118 192,113 19,075 18,088 Personal doctor 3,230 510 119 192,194 19,087 18,111 Usual source of care 3,141 496 104 189,304 18,704 17,606 Exercise 3,232 510 120 192,459 19,127 18,147 Smoking 3,226 507 120 192,052 19,061 18,110 Obesity 3,075 476 110 184,312 18,075 16,795 Diet (fewer than five servings of fruits and vegetables a day) 3,232 510 120 190,522 18,830 17,769 Alcohol intake 3,214 506 114 190,738 18,826 17,849 Fecal occult blood test 1,612 159 26* 91,400 6,824 5,614 Sigmoidoscopy 1,615 161 26* 91,035 6,714 5,612 Prostate-specific antigen test 592 65 11* 33,641 2,304 2,091 Digital rectal exam 886 95 18* 51,216 3,724 3,448 Mammogram (females) 1,362 169 31* 78,214 6,846 5,531 Clinical breast exam (females) 1,925 325 70 111,676 12,069 10,610 * We did not examine the prevalence of screening tests among Hispanics because the denominator in Delaware was less than our minimum of 50. The Delaware Department of Health and Social Services, Division of Public Health acknowledges ORC MacroSM’s analytic contributions to this report, performed under contract to the Division. QRC DIVISION OF MACRO INTERNATIONAL INC. 7315 WISCONSIN AVENUE • SUITE 400W • BETHESDA, MD 20814 Document Control Number: 35-05-20/06/07/01