Cancer is the second most common cause of death in the United States, following heart disease. The American Cancer Society projected that about 1,596,670 new cancer cases will be diagnosed and about 571,950 Americans will die of cancer in 2011 (American Cancer Society, 2011). The cancer burden varies considerably by race/ethnicity and socioeconomic status. Eliminating these disparities has been stated as an overarching objective of the Federal Healthy People 2010 initiative (U.S. Department of Health and Human Services, 2000) and the Society’s 2015 challenge goal (American Cancer Society Board of Directors, 1996). This topic reviews trends in cancer incidence, stage at diagnosis, survival, and mortality in the United States by race/ethnicity and socioeconomic status.
MEASUREMENTS AND DEFINITIONS
Incidence and mortality (or death) rates are two commonly used measurements of cancer burden. They are usually defined as the number of new cancer cases or deaths, respectively, per 100,000 persons over a specified time period. In cancer statistics, age-standardized rates are usually reported in order to minimize the effect of age when comparing rates between populations with different age structures.
Survival, a measure of disease prognosis, is the proportion of cancer patients surviving for a specified time interval after diagnosis, usually 5 years. In cancer statistics, a commonly reported survival estimate is relative survival rate, which is defined as the ratio of the observed proportion of survivors (all causes of death) among a cohort of cancer patients to the proportion of expected survivors among a comparable cohort of the general population, such that relative survival removes the effect of other causes of death. In the United States, the relative survival estimate is not available for some subpopulations, such as small racial/ethnic groups, because suitable life tables are unavailable for them. In this case, cause-specific survival, which represents a net survival of a specified disease in the absence of other causes of death, may serve as an alternative to relative survival, although cause-specific survival requires accurate classification of cause of death (Howlader et al., 2010).
The stage of cancer at the time of diagnosis is an important determinant of the choice of therapy and disease prognosis. In general, earlier stage at diagnosis is associated with better survival. In cancer registry, a summary staging scheme is applied to group invasive tumors into one of four categories: localized, regional, distant, and unstaged.
DATA COLLECTION AND REPORTING
Incidence, Stage at Diagnosis, and Survival
Since 1973, the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute (NCI) has been collecting information on the demographic characteristics of new cancer patients, extent of disease at time of diagnosis, first course of treatment, and follow-up for vital status (National Cancer Institute, 2010). Over the years, SEER has expanded from nine population-based cancer registries covering 10% of the U.S. population to 17 registries covering 28% of the population. The National Program of Cancer Registries (NPCR) (Centers for Disease Control and Prevention, 2010b) of the Centers for Disease Control and Prevention (CDC) was established by Congress through the Cancer Registries Amendment Act in 1992, and also routinely collects data on cancer occurrence, extent of disease, and initial treatment. The NPCR currently funds statewide population-based registries in 45 states, the District of Columbia, Puerto Rico, and the U.S. Pacific Island jurisdictions. The NPCR, together with SEER, currently cover the entire U.S. population.
Neither SEER nor NPCR collect individual-level socioeconomic indicators; therefore, use of SEER and NPCR data to examine disparities in cancer incidence in the United States is largely limited to analysis based on area-level indicators of socioeconomic status (SES), such as poverty rate by county, zip code, or census tract. The limitation of these ecological studies, which use aggregate data to make inferences about individuals, is that the populations within a given area may not be homogenous with respect to the particular SES variable of interest (MacRae, 1994).
Mortality
Cancer mortality data covering the entire U.S. population have been collected by the National Vital Statistics System, which is administered by the National Center for Health Statistics (NCHS) of the CDC, since 1930 (Centers for Disease Control and Prevention, 2010a). Based on information from death certificates, underlying causes of death are selected and coded according to the selection and coding rules of the revision of the International Classification of Diseases (ICD) at the time of death; the ICD is revised about every 10 years. The classification of death due to cancer is generally accurate for major cancer sites (e.g., the lung, prostate, female breast, and colorectal) (German et al., 2010; Kircher, Nelson, & Burdo, 1985). However, for some cancer sites, such as cancers in the oral cavity and pharynx, it is less consistent between the underlying cause of death and cancer registry diagnosis (German et al., 2010). NCHS has also been collecting educational attainment on death certificates, beginning in 1989 in some states and from 1993 in all states. This information has been used to describe differences in cancer mortality rates and trends by educational level (Albano et al., 2007; Kinsey, Jemal, Liff, Ward, & Thun, 2008).
RACIAL/ETHNIC DISPARITIES
In cancer statistics, cancer cases and deaths are commonly grouped into four major racial groups—White; Black, or African American (AA); Asian and Pacific Islander (API); and American Indian/Alaska Native (AI/AN)—and into two ethnic origins: non-Hispanic/Latino and Hispanic/Latino. Notably, race classification of the AI/AN group may be problematic in some cancer registries. Linkage of cancer registry data with the Indian Health Service (IHS) patient database is an effective method to reduce misclassification of the AI/AN population. However, cancer rates among AI/AN groups should still be interpreted with caution, because the IHS-linked cancer registry data are thought to lack complete representation of the AI/AN population. This linkage covers about 55% of the U.S. AI/AN population.
Incidence
Cancer incidence rates vary markedly between racial and ethnic groups (Table 2.1) (Edwards et al., 2010). For all cancers combined, Black men have a 14% higher incidence rate than White men, whereas Black women have a 7% lower incidence rate than White women (Jemal, Siegel, Xu, & Ward, 2010). For the specific cancer sites listed in Table 2.1, Blacks consistently have higher incidence rates than Whites, except for cancers of the lung and breast among women.
TABLE 2.1 Cancer Incidence Ratea by Site, Race, and Ethnicity in the United States, 2002-2006
|
|
White |
AA |
API |
AI/ANb |
Hispanic/Latinoc |
|
All sites |
|||||
|
Male |
550.1 |
626.0 |
334.5 |
441.2 |
430.3 |
|
Female |
420.0 |
389.5 |
276.3 |
369.3 |
326.8 |
|
Lung and bronchus |
|||||
|
Male |
85.9 |
104.8 |
50.6 |
78.0 |
49.2 |
|
Female |
57.1 |
50.7 |
27.6 |
56.1 |
26.5 |
|
Prostate |
146.3 |
231.9 |
82.3 |
108.8 |
131.1 |
|
Breast (female) |
123.5 |
113.0 |
81.6 |
91.7 |
90.2 |
|
Colon and rectum |
|||||
|
Male |
58.2 |
68.4 |
44.1 |
55.0 |
50.0 |
|
Female |
42.6 |
51.7 |
33.1 |
44.7 |
35.1 |
|
Kidney and renal pelvis |
|||||
|
Male |
19.7 |
20.6 |
9.0 |
24.5 |
18.2 |
|
Female |
10.3 |
10.6 |
4.5 |
15.6 |
10.3 |
|
Liver and bile duct |
|||||
|
Male |
8.0 |
12.5 |
21.4 |
12.9 |
15.9 |
|
Female |
2.8 |
3.8 |
8.1 |
6.8 |
6.2 |
|
Stomach |
|||||
|
Male |
8.9 |
16.7 |
17.5 |
14.7 |
14.3 |
|
Female |
4.2 |
8.5 |
9.8 |
7.3 |
8.6 |
|
Uterine cervix |
7.9 |
11.1 |
7.6 |
9.4 |
12.7 |
Incidence rates were calculated based on the registry data in 43 states.
AA, African American; API, Asian American and Pacific Islander; AI/AN, American Indian and Alaska Native.
aPer 100,000, age adjusted to the 2000 U.S. standard population.
bData based on Contract Health Service Delivery Areas, comprising about 55% of the U.S. American
Indian/Alaska Native population.
cPersons of Hispanic/Latino origin may be of any race.
The higher breast cancer incidence rates among White women may be attributable to the combined effect of cancer screening (e.g., more frequent mammography in White women) and exposure to risk factors (e.g., later age at first birth and greater use of menopausal hormone therapy among White compared to Black women) (Ghafoor et al., 2003).
Compared to Whites and Blacks, other racial/ethnic groups have a lower incidence rate for all cancers combined (Table 2.1) . With respect to site-specific cancer incidence, these minority groups have a lower rate than Whites for most of the common cancer sites, but typically have a higher rate for cancer sites related to infectious agents, such as the uterine cervix, liver, and stomach. For example, the incidence rate for cancer of the uterine cervix is 60% higher among Hispanics/Latinos than among Whites. Asian American/Pacific Islanders have the highest incidence rates for both liver and stomach cancers in both men and women. Higher stomach and liver cancer incidence rates in the API population may reflect an increased prevalence of chronic infection with Helicobacter pylori and hepatitis B and C viruses, respectively, in this population (Ward et al., 2004). American Indians and Alaska Natives have a higher incidence rate for kidney cancer than other racial/ethnic group. The higher prevalence of obesity and smoking among American Indians and Alaska Natives may contribute to this disparity (Espey et al., 2007).
Although cancer incidence varies largely across racial/ethnic groups, prostate, lung, and colorectal cancer in men, and breast, lung, and colorectal cancer in women, without specific rank order, are the three most commonly diagnosed cancers in each racial/ethnic group (Edwards et al., 2010).
Stage at Diagnosis and Survival
African Americans and other minority groups are more likely than Whites to be diagnosed at a later stage of disease for all the four major cancers: lung, colorectal, female breast, and prostate cancers (Figure 2.1) (Altekruse et al., 2010). For lung cancer, the proportion of distant stage diagnoses is 2%-8 % points higher in minority groups than in Whites; for breast cancer, the percentage of localized stage diagnoses is 10 % points higher in White women than in Black women.
Since 1975, survival rates for the four common cancer sites, with the exception of the lung, have substantially improved among both African Americans and Whites (Table 2.2) (Altekruse et al., 2010). However, African American men and women continue to have poorer survival than their White counterparts for nearly every cancer site. During the period from 1999 to 2006, the relative 5-year survival rate in African Americans was at least 10 % points lower than in Whites for all cancers combined (59.2% in Blacks vs. 69.1% in Whites) and cancers of the colorectum (56.7% in Blacks vs. 67.9% in Whites) and female breast (78.4% in Blacks vs. 91.2% in Whites). Because accurate life expectancies are not available for racial and ethnic groups other than Whites and Blacks, relative survival rates cannot be calculated for those minority populations. However, an analysis of cause-specific survival rates among cancer patients diagnosed between 1999 and 2006 in 17 SEER areas reported that all minority male populations (Blacks, Asians/Pacific Islanders, American Indians/Alaska Natives, and Hispanics) had a greater probability of dying from cancer within 5 years of diagnosis than White men. Among women, Asians/Pacific Islanders have the highest (68.2%) and African Americans have the lowest 5-year cancer-specific survival rate (56.0%) (Table 2.3) (Altekruse et al., 2010).
FIGURE 2.1 The Distribution of Selected Cancers by Race/Ethnicity and Stage at Diagnosis, United States, 1999-2006
For each cancer type, stage categories do not total 100% because insufficient information is available to assign a stage to all cancer cases.
Mortality
For all cancers combined, Black men have a 34% higher death rate than White men; and Black women have a 17% higher death rate than White women (Jemal et al., 2010). For the specific cancer sites listed in Table 2.4, African Americans consistently have higher death rates than Whites, except for lung cancer among women and kidney cancer among both men and women.
Racial/ethnic groups other than Blacks have a lower death rate than Whites for all cancers combined and for most of the common cancer sites. However, they have higher death rates for cancers of the liver, kidney, and uterine cervix. For example, death rates for both liver and stomach cancer among Asian Americans and Pacific Islanders are almost twice as high as those in Whites. The death rate from kidney cancer in American Indians and Alaska Natives is about 1.5 times as high as that in Whites.
Similar to the incidence data, lung, prostate, and colorectal cancer in men and lung, breast, and colorectal cancer in women, are the three leading causes of cancer death in all racial/ethnic groups, except in API men, in which liver cancer, but not prostate cancer, is among the three leading causes of cancer death (Edwards et al., 2010).
TABLE 2.2 Five-Year Relative Survival Rates (%) by Race and Year of Diagnosis, United States, 1975-2006
|
All Cancers Combined |
Lung Cancer |
Colorectal Cancer |
Breast Cancer (Female) |
Prostate Cancer |
||||||
|
White |
Black |
White |
Black |
White |
Black |
White |
Black |
White |
Black |
|
|
1975-1977 |
51.2 |
39.9 |
12.8 |
11.6 |
51.2 |
46.3 |
76.1 |
62.4 |
70.2 |
61.4 |
|
1978-1980 |
51.3 |
39.6 |
13.4 |
12.0 |
52.7 |
46.2 |
75.5 |
63.9 |
72.9 |
62.8 |
|
1981-1983 |
52.6 |
39.6 |
13.9 |
11.7 |
55.6 |
47.5 |
77.6 |
64.1 |
74.6 |
63.7 |
|
1984-1986 |
54.8 |
40.8 |
13.6 |
11.4 |
59.4 |
49.2 |
80.5 |
65.3 |
77.6 |
66.3 |
|
1987-1989 |
57.8 |
43.6 |
13.8 |
11.2 |
61.1 |
53.3 |
85.4 |
71.3 |
85.3 |
72.1 |
|
1990-1992 |
62.3 |
48.1 |
14.4 |
10.7 |
63.0 |
53.6 |
86.7 |
71.7 |
94.8 |
85.2 |
|
1993-1995 |
63.6 |
53.1 |
15.1 |
13.2 |
61.7 |
52.9 |
88.0 |
72.8 |
96.6 |
92.2 |
|
1996-1998 |
65.8 |
55.7 |
15.5 |
12.7 |
64.3 |
55.1 |
89.7 |
76.4 |
98.6 |
95.2 |
|
1999-2006 |
69.1 |
59.2 |
16.8 |
13.2 |
67.9 |
56.7 |
91.2 |
78.4 |
99.9 |
97.3 |
|
Absolute change |
17.9 |
19.3 |
4.0 |
1.6 |
16.7 |
10.4 |
15.1 |
16.0 |
29.7 |
35.9 |
|
Relative change3 |
35.0 |
48.4 |
31.3 |
13.8 |
32.6 |
22.5 |
19.8 |
25.6 |
42.3 |
58.5 |
aRelative change = absolute difference in 5-year relative survival rates between 1975-1977 and 1999-2006, as a percentage of the rate of 1975-1977.
TABLE 2.3 Five-Year Cause-Specific Survivala (%) for All Cancers Combined and Selected Cancers, 1999-2006
|
|
White |
AA |
API |
AI/ANb |
Hispanic/Latinoc |
|
All cancers combined |
|||||
|
Male |
65.6 |
60.9* |
59.8* |
53.5* |
63.3* |
|
Female |
65.6 |
56.0* |
68.2* |
60.6* |
65.5 |
|
Lung cancer |
|||||
|
Male |
15.0 |
12.5* |
16.0 |
12.0 |
13.6* |
|
Female |
19.7 |
16.1* |
22.1* |
12.5* |
17.2* |
|
Colorectal cancer |
|||||
|
Male |
64.9 |
55.7* |
67.5* |
59.0* |
61.6* |
|
Female |
63.7 |
56.7* |
68.6* |
65.5 |
62.4 |
|
Breast cancer (female) |
88.3 |
77.5* |
90.3* |
84.6* |
85.3* |
|
Prostate cancer |
93.5 |
91.2* |
93.8 |
89.7* |
92.3* |
AA, African American; API, Asian American and Pacific Islander; AI/AN, American Indian and Alaska Native.
aFive-Year Cause-Specific Survival was calculated by actuarial method using diagnosis years 1999-2006, with follow-up through 2007.
bData based on Contract Health Service Delivery Areas. cPersons of Hispanic/Latino origin may be of any race.
^Statistically significantly different from the rate of Whites (p < .05).
It is important to note that the cancer disparities described earlier are based on broad racial/ethnic groups that are not homogenous; thus, differences in the cancer burden also exist within each racial/ethnic group. For example, among Asian Americans and Pacific Islanders, native Hawaiians and Samoans have the highest overall cancer incidence and death rates, which are two to three times those for Asian Indians; Chinese and Vietnamese have the highest incidence of and death rates from nasopharyngeal cancer; Vietnamese women are three times as likely to be diagnosed with cervical cancer as Chinese and Japanese women; and Laotians and Samoans have lower percentages of early-stage cancers of the colorectum, breast (female), and cervix uteri, compared with other API groups (American Cancer Society, 2010; Miller, Chu, Hankey, & Ries, 2008).
Racial/ethnic cancer disparities are thought to result from a complex interplay of numerous factors affecting cancer occurrence and outcomes. Minority groups often face obstacles to receiving health care services related to cancer prevention, early detection, and high-quality treatment because of low income, inadequate health insurance, cultural and language barriers, and racial bias. According to data from the National Health Interview Survey (NHIS) in 2009, poverty rates and the likelihood of being medically uninsured are approximately two to three times higher for African Americans and Hispanics/Latinos than for non-Hispanic Whites (Cohen, Martinez, & Ward, 2010). In addition, minority groups often have lower education levels, which are associated with a higher prevalence of risk factors for cancer (Centers for Disease Control and Prevention, 2009, 2010c) and less access to timely or high-quality treatment (Berry et al., 2009; Rolnick et al., 2005). Overall, these socioeconomic disadvantages lead to increased risk of cancer occurrence, more advanced stage at cancer diagnosis, and poorer survival among minority groups, especially African Americans. In addition to social factors, genetic variations are also thought to contribute to the disproportionate cancer burden between racial/ethnic groups; however, genetic differences are thought to make only a minor contribution to cancer disparities (Hemminki, Forsti, & Lorenzo Bermejo, 2008). Detailed descriptions of disparities in cancer risk factors, screening, treatment, and genetic variations can be found in other topics.
TABLE 2.4 Cancer Death Ratea by Site, Race, and Ethnicity in the United States, 2002-2006
|
|
White |
AA |
API |
AI/ANb |
Hispanic/Latinoc |
|
All sites |
|||||
|
Male |
226.7 |
304.2 |
135.4 |
183.3 |
154.8 |
|
Female |
157.3 |
183.7 |
95.1 |
140.1 |
103.9 |
|
Lung and bronchus |
|||||
|
Male |
69.9 |
90.1 |
36.9 |
48.0 |
33.9 |
|
Female |
41.9 |
40.0 |
18.2 |
33.5 |
14.4 |
|
Prostate |
23.6 |
56.3 |
10.6 |
20.0 |
19.6 |
|
Breast (female) |
23.9 |
33.0 |
12.5 |
17.6 |
15.5 |
|
Colon and rectum |
|||||
|
Male |
21.4 |
31.4 |
13.8 |
20.0 |
16.1 |
|
Female |
1 4.9 |
21.6 |
10.0 |
13.7 |
10.7 |
|
Kidney and renal pelvis |
|||||
|
Male |
6.1 |
6.0 |
2.4 |
9.0 |
5.2 |
|
Female |
2.8 |
2.7 |
1.2 |
4.2 |
2.4 |
|
Liver and bile duct |
|||||
|
Male |
6.8 |
10.8 |
15.0 |
10.3 |
11.3 |
|
Female |
2.9 |
3.9 |
6.6 |
6.5 |
5.1 |
|
Stomach |
|||||
|
Male |
4.8 |
11.0 |
9.6 |
9.8 |
8.3 |
|
Female |
2.4 |
5.3 |
5.8 |
4.6 |
4.8 |
|
Uterine cervix |
2.2 |
4.6 |
2.2 |
3.4 |
3.1 |
AA, African American; API, Asian American and Pacific Islander; AI/AN, American Indian and Alaska Native.
aPer 100,000, age adjusted to the 2000 U.S. standard population.
bData based on Contract Health Service Delivery Areas, comprising about 55% of the U.S. American Indian/Alaska Native population.
cPersons of Hispanic/Latino origin may be of any race.
