Fax: (770) 488-4639
An updated analysis of data from the National Program of Cancer Registries and the Surveillance, Epidemiology, and End Results Program
Article first published online: 20 FEB 2009
Published 2009 American Cancer Society
Volume 115, Issue 9, pages 1967–1976, 1 May 2009
How to Cite
Rim, S. H., Seeff, L., Ahmed, F., King, J. B. and Coughlin, S. S. (2009), Colorectal cancer incidence in the United States, 1999-2004 . Cancer, 115: 1967–1976. doi: 10.1002/cncr.24216
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
This article is a US government work and, as such, is in the public domain in the United States of America.
- Issue published online: 20 APR 2009
- Article first published online: 20 FEB 2009
- Manuscript Accepted: 27 OCT 2008
- Manuscript Revised: 20 OCT 2008
- Manuscript Received: 26 JUN 2008
- colorectal cancer;
- National Program of Cancer Registries;
- End Results
By using recent national cancer surveillance data, the authors investigated colorectal cancer (CRC) incidence by subpopulation to inform the discussion of demographic-based CRC guidelines.
Data included CRC incidence (1999-2004) from the combined National Program of Cancer Registries and Surveillance, Epidemiology, and End Results Program databases. Incidence rates (age-specific and age-adjusted to the 2000 US standard population) were reported among individuals ages 40 to 44 years, 45 to 49 years, 50 to 64 years, and ≥65 years by sex, subsite, disease stage, race, and ethnicity. Rate ratios (RR) and rate differences (RD) were calculated to compare CRC rates in different subpopulations.
Incidence rates were greater among men compared with women and among blacks compared with whites and other races. Incidence rates among Asians/Pacific Islanders (APIs), American Indians/Alaska Natives (AI/ANs), and Hispanics consistently were lower than among whites and non-Hispanics. Sex disparities were greatest in the population aged ≥65 years, whereas racial disparities were more pronounced in the population aged <65 years. Although the RD between blacks and whites diminished at older ages, the RD between APIs and whites, between AI/ANs and whites, and between non-Hispanics and Hispanics increased with increasing age. By subsite, blacks had the highest incidence rates compared with whites and other races in the proximal and distal colon; the reverse was true in the rectum. By stage, whites had higher incidence rates than blacks and other races for localized and regional disease; for distant and unstaged disease, blacks had higher incidence rates than whites.
The current findings suggested differences that can be considered in formulating targeted screening and other public health strategies to reduce disparities in CRC incidence in the United States. Cancer 2009. Published 2009 by the American Cancer Society.
Colorectal cancer (CRC) is the third most commonly diagnosed cancer and the second leading cause of cancer deaths in the United States among cancers that affect both men and women.1, 2 Although CRC incidence and mortality rates have declined in recent years for men and women of all races/ethnicities,1 the burden of disease remains high and disproportionate within demographic subpopulations.3
A substantial amount of evidence has demonstrated that regular screening for CRC is effective in reducing incidence and mortality from this disease.4-9 On the basis of these data, several national organizations, including the US Preventive Services Task Force, the American Cancer Society, and the US Multi-Society Task Force on CRC Screening, have strongly recommended CRC screening for all average risk men and women aged ≥50 years regardless of sex, race, or ethnicity with 1 of the following screening modalities: fecal occult blood test (FOBT) annually, flexible sigmoidoscopy every 5 years, a combination of FOBT and flexible sigmoidoscopy, double-contrast barium enema every 5 years, and colonoscopy every 10 years or as a follow-up examination.10-12
Recently, however, several organizations have begun to prescribe CRC screening guidelines to specific demographic subpopulations. In 2005, the American College of Gastroenterology (ACG) issued recommendations for CRC screening that were specific for African Americans based on literature indicating an increased disease burden in this population.13 Since then, many organizations have updated their guidelines to take positions on a preferred screening modality (American College of Obstetricians and Gynecologists), to include more screening tests (American Cancer Society), and/or to put an upper age limit to recommended routine CRC screening (US Preventive Services Task Force). The debate about tailoring guidelines to CRC burden in demographic subgroups continues.
To describe the magnitude of disparities in CRC disease burden and to inform the discussion of burden-based guidelines and other prevention strategies, we used recent, nationally representative cancer surveillance data to describe the incidence of invasive CRC in the United States from 1999 to 2004 by sex, race, ethnicity, age at diagnosis, anatomic subsite, and stage at diagnosis. Our objective was 2-fold: to describe the most current US CRC incidence data and to look specifically at burden of disease at the proposed younger age cutoff.
MATERIALS AND METHODS
Data Source and Study Population
CRC incidence data were obtained from US population-based cancer registries participating in the National Program of Cancer Registries (NPCR) of the Centers for Disease Control and Prevention (CDC) and/or the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute (NCI).2, 14 The SEER Program began collecting data in specific geographic regions in 1973. The NPCR was established in 1992 in response to the congressional Cancer Registries Amendment Act (Public Law 102-515).15 Data from both the NPCR and the SEER Program are reported using standardized data items and codes according to the North American Association of Central Cancer Registries.16 Together, these registries cover 100% of the US population. However, because we excluded registries that did not meet high-quality data criteria2 for all study years, 1999 through 2004, our analytic file covered 92% of the US population and represented 44 states and the District of Columbia. Data for the population denominators that were used to calculate incidence rates were derived from the 2000 US Census.2 The data used for this study were reported to the CDC as of January 31, 2007 and to the NCI as of November 1, 2006. The NPCR Cancer Surveillance System was approved for research (protocol 2594) by the CDC Institutional Review Board.
Definition of Variables
Incidence rates were stratified by age, sex, race/ethnicity, anatomic subsite, and stage at diagnosis. We analyzed data in 5-year age groups (ages 40-44 years, 45-49 years, 50-64 years, and ≥65 years). Individuals aged <40 years were excluded from the analysis because of small overall case counts for the combined years. We used these age categories to assess the burden of disease in individuals for whom early screening has been recommended (aged ≥45 years) and those for whom screening currently is recommended (aged ≥50 years), for the pre-Medicare and Medicare-eligible populations. Racial groups were defined as white, African American, Asian or Pacific Islander (API), American Indian/Alaska Native (AI/AN), and “all races,” which also included individuals of other or unknown race. Race and ethnicity categories were not mutually exclusive. Ethnicity was defined as Hispanic and non-Hispanic.
The anatomic subsites for proximal colon, distal colon, and rectum were categorized according to International Classification of Diseases for Oncology, third edition (ICD-0-3) topography codes and were grouped as noted by Wu et al.17 The right or proximal colon included cancers of the cecum (ICD-0-3 code C18.0), ascending colon (code C18.2), hepatic flexure (code C18.3), transverse colon (code C18.4), and splenic flexure (code C18.5). The proximal colon included subsites that most commonly are evaluated by colonoscopy only and may not be within the reach of a sigmoidoscope. The left or distal colon included the descending colon (code C18.6) and the sigmoid colon (code C18.7) or cancers within the reach of a flexible sigmoidoscope. The rectum included the rectosigmoid junction (code C19.9) and the rectum, not otherwise specified (code C20.9). Statistical analyses for “all colorectal cancers” also included cancers of the appendix (code C18.1); the colon, not otherwise specified (code C18.8, 18.9); and the intestinal tract, not otherwise specified (code C26.0). Although there are several ways to define the proximal versus distal sites, we chose this definition because it aligns better with the use of colonoscopy versus sigmoidoscopy.
Categories for the stage at diagnosis followed SEER Summary Stage guidelines.18 Malignant, microscopically confirmed tumors were defined as localized (invasive tumors confined to the colon or rectum), regional (tumors that invaded surrounding tissues, organs, lymph nodes), distant (metastasized tumors), and unstaged or unknown disease.18
We described incidence by US Census regions. Coverage of the study population was 100% for the Northeast and West, 98% for the Midwest, and 80% for the South.
Data were analyzed using SEER*Stat software (version 6.3.5),14 and incidence rates of invasive cancers were reported per 100,000 population. Incidence rates were age-adjusted to the 2000 US standard population using the direct method; the 95% confidence intervals of the age-adjusted rates were calculated using the method described by Tiwari et al.19
Measures of cancer disparities were consistent with previous literature20-24 indicating the importance of considering disparities in both relative and absolute terms to gain a better understanding of their magnitude. Although there are several ways to measure health disparities, we chose 2 frequently used relative and absolute measures, rate ratios (RRs) and rate differences (RD), respectively, to describe the excess CRC burden in specific subpopulations. Rate comparisons were expressed as RRs and RDs.22 The RR was defined as the ratio of age-adjusted incidence rates (and sometimes is reported in percentages); the RD was calculated as the difference between age-adjusted incidence rates (per 100,000) in a specific population compared with a reference population. Referent groups, whites for race and non-Hispanics for ethnicity, were selected based on relative size, with the largest group having more stable rates. Statistical significance was assessed at P < .05 (2-tailed). Statistics were not calculated for population subgroups that had <16 cases.
During the combined years from 1999 to 2004, in total, 875,908 eligible CRCs were reported in the United States, of which 7% were in situ tumors and 93% were invasive, malignant tumors. For this analysis, we considered invasive tumors only (n = 814,697) (Table 1). Within this study population, 51% of invasive CRCs (415,278) occurred in men, and 49% (399,419) occurred in women. By age group, 92% of invasive CRCs occurred among individuals aged ≥50 years, and 67% occurred among individuals aged ≥65 years. By subsite, 42% of invasive CRCs were located in the proximal colon. By stage, 37% of invasive CRCs were diagnosed as localized disease, 55% were diagnosed at later stage (regional or distant) disease, and 7% were unstaged.
|Characteristic||No. of Cases (%)||Age-Adjusted Rate [95% CI]†|
|Overall (US)||814,697||51.5 [51.4-51.6]|
|Men||415,278 (51.0)||61.1 [60.9-61.3]|
|Women||399,419 (49.0)||44.2 [44.1-44.4]|
|All races‡||814,697||51.5 [51.4-51.6]|
|White||705,985 (86.7)||50.8 [50.7-50.9]|
|Black||77,839 (9.6)||57.2 [56.8-57.6]|
|API||19,584 (2.4)||38.9 [38.3-39.4]|
|AI/AN||3073 (0.4)||31.3 [30.1-32.4]|
|Non-Hispanic||770,673 (94.6)||52.2 [52.1-52.3]|
|Hispanic§||44,024 (5.4)||41.5 [41.1-41.9]|
|0-39||17,381 (2.1)||2.0 [1.9-2.0]|
|40-44||18,722 (2.3)||14.9 [14.7-15.1]|
|45-49||32,800 (4.0)||28.4 [28.1-28.7]|
|50-64||199,371 (24.5)||81.2 [80.8-81.5]|
|≥65||546,423 (67.1)||276.5 [275.8-277.2]|
|Stage at diagnosis|
|Localized||303,260 (37.2)||19.2 [19.1-19.3]|
|Regional||316,405 (38.8)||20 [19.9-20.1]|
|Distant||135,344 (16.6)||8.6 [8.5-8.6]|
|Unstaged||59,682 (7.3)||3.8 [3.7-3.8]|
|Proximal colon||344,446 (42.3)||21.8 [21.7-21.9]|
|Distal colon||205,973 (25.3)||13.0 [13.0-13.1]|
|Rectum||223,400 (27.4)||14.1 [14.0-14.2]|
|Northeast||198,046 (24.3)||56.1 [55.9-56.4]|
|Midwest||209,005 (25.7)||53.6 [53.4-53.9]|
|South||246,193 (30.2)||50.6 [50.4-50.8]|
|West||161,453 (19.8)||45.8 [45.6-46.1]|
The overall age-adjusted incidence rate of invasive CRC incidence in the United States during the study period was 51.5 per 100,000 (95% confidence interval, 51.4-51.6 per 100,000) for all individuals and all ages (Table 1); incidence was higher for men compared with women (61.1 per 100,000 men vs 44.2 per 100,000 women). By race, the incidence was highest among blacks (57.2 per 100,000) compared with whites (50.8 per 100,000) and APIs (38.9 per 100,000). Incidence rates increased with increasing age. The majority of invasive cancers were detected at the localized or regional stage. More left-sided (distal plus rectal) than right-sided (proximal) cancers were detected during the study period; however, among the 3 subsites, incidence rates were highest for proximal cancers. Rates of CRC were highest in the Northeast and Midwest compared with other Census regions of the United States.
The greatest difference in rates by sex was among men and women aged ≥50 years; incidence rates were >40% higher for men than for women (among individuals aged 50-64 years, the respective rates were 96.2 vs 67.2 per 100,000; among individuals aged ≥65 years, the respective rates were 332.6 vs 237.3 per 100,000) (Table 2). In absolute terms, there were 29 additional invasive CRCs per 100,000 men compared with women ages 50 to 64 years (RD, 29 per 100,000), and 95 more cancers per 100,000 men compared with women aged ≥65 years (RD, 95.3 per 100,000).
|Age, y||No.||Rate (95% CI)†||No.||Rate (95% CI)†||RR‡||RD§|
|40-44||9659||15.5 (15.2-15.8)||9063||14.4 (14.1-14.7)||1.08∥||1.1|
|45-49||17,342||30.4 (30.0-30.9)||15,458||26.4 (26-26.8)||1.15∥||4.0|
|50-64||113,918||96.2 (95.7-96.8)||85,455||67.2 (66.8-67.7)||1.43∥||29.0|
|≥65||265,351||332.6 (331.3-333.9)||281,072||237.3 (236.4-238.2)||1.40∥||95.3|
|All ages¶||406,270||139.3 (138.8-139.7)||391,048||100.3 (100-100.6)||1.39∥||39.0|
Blacks had significantly higher incidence rates than whites, APIs, and AI/ANs in all age groups (Table 3). This was most evident in the population aged <65 years. In the groups ages 45 to 49 years and 50 to 64 years, blacks had incidence rates that were 40% and 35% higher than whites, respectively. In absolute terms, blacks had 10 more cases of invasive cancers per 100,000 than whites in the age group 45 to 49 years and 27 more cases of invasive cancers per 100,000 than whites in the age group 50 to 64 years. In all age groups, APIs and AI/ANs had lower incidence rates than both blacks and whites, and Hispanics had lower incidence rates than non-Hispanics. The RD between APIs and whites, between AI/ANs and whites, and between non-Hispanics and Hispanics increased with increasing age.
|Age Group and Race/ Ethnicity||No.||Age-Adjusted Rate (95% CI)†||RR‡||RD§|
|Ages 40-44 y|
|All races∥||18,722||14.9 (14.7-15.1)|
|Ages 45-49 y|
|All races∥||32,800||28.4 (28.1-28.7)|
|Ages 50-64 y|
|All races∥||199,371||81.2 (80.8-81.5)|
|Ages ≥65 y|
|All races∥||546,423||276.5 (275.8-277.2)|
|All races∥||797,316||117.1 (116.8-117.4)|
Incidence Rates by Subsite
Figure 1 illustrates invasive CRC incidence rates by subsite, age group, race, and ethnicity. The greatest burden of disease was in the proximal colon, where case counts and age-adjusted incidence rates were higher (21.8 per 100,000 population) compared with the distal colon (13 per 100,000 population) and the rectum (14.1 per 100,000 population) (Table 1). Incidence rates in the proximal colon were highest in the population aged ≥65 years and among blacks compared with other races. In the distal colon, rates also were higher among blacks, although the RD was smaller between groups. Although the relative rate increase by race was greater in the age group 45 to 49 years for men and women, the absolute rate increase was greater at older ages. Incidence rates among APIs were lower than the rates among whites and blacks in the proximal and distal colon for all age groups. CRC incidence rates were lowest among AI/ANs in all subgroups except those ages 50 to 64 years with proximal cancers, for which the AI/AN population had slightly higher incidence rates than APIs. In the rectum, rates did not differ substantially by race, except among those aged ≥65 years; in that group, whites had higher incidence rates compared with all other racial groups.
Incidence Rates by Stage
Figure 2 illustrates the incidence of invasive CRC by stage, age group, race, and ethnicity. Among individuals aged <65 years, blacks had higher incidence rates than all other race groups at all stages. Among individuals aged ≥65 years, blacks had a lower incidence of localized disease than whites but had higher incidence rates of distant and unstaged cancers. Regional stage disease revealed nearly no difference according to race (black vs white) among individuals aged ≥65 years. For APIs, incidence rates were lower overall than the rates among whites for all age groups and stages. Similarly for AI/ANs, incidence rates were lower than all other racial/ethnic groups of all ages and stages; however, at distant stage, there was little variation by race, particularly among APIs, AI/ANs, and Hispanics. The magnitude of decrease in age-adjusted incidence rates for each stage differed by age. Hispanics had lower incidence rates than non-Hispanics for all stages except unstaged disease.
To our knowledge, the current study provides the most current national CRC incidence rates in the United States by subpopulations. With combined NPCR and SEER data, we were able to provide a comprehensive picture of the burden of disease by sex, age, race/ethnicity, subsite and stage. By using RRs and RDs, we evaluated the magnitude of difference in incidence between demographic subgroups.
Although national CRC incidence rates have decreased slightly compared with previous studies of incidence data from the late 1990s and early 2000s, disparities by sex and race persist.3 Over the 6-year period of the current analysis, incidence rates were higher among men than women, higher among blacks than other races, and higher among non-Hispanics than Hispanics. Incidence rates increased with increasing age, particularly after age 50 years. Most cancers were detected at localized and regional stages and were located more in the proximal colon than the distal colon or rectum.
Men had higher CRC incidence rates than women in all age groups and at all subsites of the colon, and the sex disparity increased with age. Subsite distribution was similar for men and women. Blacks had higher CRC incidence rates than whites or APIs in all age groups. This disparity by race was most evident in the population aged <65 years, whereas incidence rates by race essentially were the same in the population aged ≥65 years.
Blacks in all age groups had the highest incidence rates for cancers in the proximal and distal colon, whereas older whites had the highest rates of rectal cancers. Localized and regional cancers were diagnosed most commonly among whites and non-Hispanics and were diagnosed least commonly among blacks in all age groups. Conversely, distant or unstaged cancers were diagnosed more frequently in blacks. For all individuals, CRCs occurred most commonly in the proximal colon compared with other subsites, and highest rates of proximal cancer were observed in the population aged ≥65 years. Many of these patterns in CRC incidence were observed in previous analyses.1-3, 17, 25-30
These disparities in CRC incidence rates by sex and race likely exist from a complex interplay of genetic, hormone, and environmental risk factors and from the variability in population-level screening rates and access to care.31-37 Access to care may be limited because of lower income, residence in socioeconomically distressed counties or in rural/medically underserved urban areas, or lack of healthcare insurance or under-insurance.38, 39 Factors that may be associated with the risk of developing CRC, including physical inactivity, obesity, alcohol use, and diabetes, among other environmental and host factors, also may vary by demographic subgroup and may play a role in these patterns of cancer incidence.40
It is difficult to tease out the relative influences of biology, access to care, and other potential factors on these incidence patterns. For example, both the younger age at presentation and the subsite distribution of CRCs in blacks compared with other races may suggest a difference in biology of the disease by race or a common set of environmental influences.27 However, the difference in stage by race suggests that there are disparities in access to care and screening.41, 42 Previous studies have identified several factors associated with late stage CRC diagnosis, including lack of adherence to guidelines for screening, age, less education, race, Hispanic ethnicity, and factors associated with decreased access to care.28, 37, 39, 41-43
It has been demonstrated that access to care and screening varies by race, income, and insurance status.39, 44 Although screening rates for the US population overall are increasing, the rates remain lower for blacks and for individuals with low income or without health insurance.44 In addition, the type of screening tests received varies by sex and race. Previous reports of national CRC screening test prevalence have indicated that more men report being screened with endoscopy than women, whereas more women are screened with FOBT; and, by race, whites are more likely to have reported being screened with endoscopy than blacks.37 Endoscopic screening tests have a higher sensitivity for polyps and cancers than FOBT, which also may be contributing to the incidence patterns by sex and race. Furthermore, until recently, sigmoidoscopy, a test that provides a direct visual view of only the distal colon and rectum, was used more commonly as an endoscopic screening test than colonoscopy; the more prevalent use of endoscopic tests by whites than blacks may explain in part the difference in subsite distribution by race.
Our data can offer some insight into the discussion regarding whether screening strategies should be based on disease burden. Currently, most national CRC guidelines do not make screening recommendations that differ by sex or race.45, 46 However, some national organizations have revisited guidelines to promote the use of colonoscopy among women47; screening tests that detect precancerous polyps as well as cancer45; and, most recently, recommendations against routine CRC screening to average-risk age groups aged >75 years.46
To date, 1 organization has taken the initiative of considering tailoring guidelines by race. In 2005, the ACG recommended that African Americans should begin screening at age 45 years using colonoscopy as the preferred, initial screening test.13 That recommendation was based on older incidence data demonstrating the occurrence of more right-sided disease in younger African Americans.13 Our results confirm these racial disparities. We demonstrated that, in the group ages 45 to 49 years, blacks had a 40% higher incidence rate than whites, with 10 additional cancers per 100,000 (or 1 in 10,000) among blacks compared with whites. Moreover, the results suggest that screening with colonoscopy among blacks ages 45 to 49 years would detect approximately 6 more (proximal) cancers per 100,000 than whites that otherwise would have been missed by flexible sigmoidoscopy or a different screening modality. In interpreting these results, it is important to be mindful that the relative risk can be inconsistent with the underlying absolute difference, leading to different conclusions.22, 23
If screening guidelines were lowered to age 45 years, then efforts to detect these additional cancers should be considered in the context of several other factors, including unintended complications incurred from screening, the cost of screening and resource allocation, the provider or health system capacity to screen additional individuals, and the potential impact of early screening on population incidence and mortality. Measures like efficiency frontier, to help define age parameters for which the greatest numbers of life years gained are balanced with resource expenditures, also may help guide these decisions.48 In addition, potential risks and benefits both to the population and to each individual can be weighed when preventive interventions are being considered.49 Finally, as the discussion of “customized” guidelines continues, the broader implications of these potential changes also should be considered. For example, if screening should be initiated earlier in African Americans because of their earlier onset of disease, then should subgroups with lower incidence rates also have tailored guidelines? What benchmarks should be used in deciding these cutoff ages?
Although the combined NPCR and SEER dataset represents high-quality data with national population-based coverage, there are some inherent data limitations.15 These data are descriptive and do not allow for any assessment of causality. For our analysis, because of small sample sizes in certain subgroups, we did not create racial groups that were mutually exclusive of Hispanic ethnicity. We also did not report incidence rates among AI/ANs. Finally, although NPCR/SEER data capture the national majority,2 our dataset only included states that met high-quality data criteria and represented 92% of the US population from 1999 to 2004. Thus, we have under-report the total number of cases in the United States and are unable to know whether there were different patterns in populations that were not included in this analysis.
All in all, our findings indicated the greatest disparities by race (blacks vs whites) in the population aged ≤65 years, whereas sex disparities were greatest in the population aged ≥65 years. These relative and absolute differences in incidence rates by subpopulation can be considered in formulating screening and other population-based strategies for reducing the CRC burden. While the discussion of demographic-based guidelines progresses, persistent patterns of disparities in cancer incidence should be addressed comprehensively, with caution, and taking into consideration the perspectives at the provider, patient, and public health levels on the cost of screening, resource allocation, cultural/historic barriers, and other issues that may contribute to disparities.
Conflict of Interest Disclosures
The authors made no disclosures.
- 2U.S. Cancer Statistics Working Group. United States Cancer Statistics: 2004 Incidence and Mortality. Atlanta, Ga: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, and National Cancer Institute; 2007.
- 13Colorectal cancer in African Americans. Am J Gastroenterol. 2005; 100: 515-523., , , et al.Direct Link:
- 16North American Association of Central Cancer Registries (NAACCR). NAACCR Data Standards for Cancer Registries. Springfield, Ill: NAACCR; 2008; Available at: http://www.naaccr.org. Accessed on June 26, 2008.
- 18Young JJ, Roffers S, Ries L, Fritz A, Hurlbut A, eds. SEER Summary Staging Manual-2000: Codes and Coding Instructions. NIH Publication No. 01-4969. Bethesda, Md: National Cancer Institute; 2001.
- 22Methodological Issues in Measuring Health Disparities. Report No. 141. Hyattsville, Md: National Center for Health Statistics; 2005., , , et al.
- 23Methods for Measuring Cancer Disparities: Using Data Relevant to Healthy People 2010 Cancer-Related Objectives. NCI Cancer Surveillance Monograph Series, No. 6. NIH Publication 05-5777. Bethesda, Md: National Cancer Institute; 2005., .
- 40Cancers of the colon and rectum. In: Schottenfeld D, Fraumeni J, eds. Cancer Epidemiology and Prevention.3rd ed. New York, NY: Oxford University Press; 2006: 809-829., .
- 44Centers for Disease Control and Prevention. Use of colorectal cancer tests—United States, 2002, 2004, and 2006. MMWR Morb Mortal Wkly Rep. 2008; 57: 253-258.
- 45Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. CA Cancer J Clin. 2008; 58: 130-160., , , et al.
- 48Technology Assessments for Cost-Effectiveness of DNA Stool Test to Screen for Colorectal Cancer. Washington, DC: Department of Health and Human Services, Centers for Medicare & Medicaid Services; 2007. Available at: http://www.cms.hhs.gov/mcd/viewtechassess.asp?from2=viewtechassess.asp&id=212& Accessed on October 19, 2008., , , , , .