Clinical outcomes from the CDC's Colorectal Cancer Screening Demonstration Program


  • The articles in this supplement were commissioned based on participation in evaluating the Centers for Disease Control and Prevention-funded Colorectal Cancer Screening Demonstration Program.

  • The opinions or views expressed in this supplement are those of the authors and do not necessarily reflect the opinions or recommendations of the journal editors, the American Cancer Society, John Wiley & Sons, Inc., or the Centers for Disease Control and Prevention.



Colorectal cancer remains the second leading cause of cancer-related deaths among US men and women. Screening rates have been slow to increase, and disparities in screening remain.


To address the disparity in screening for this high burden but largely preventable disease, the Centers for Disease Control and Prevention (CDC) designed and established a 4-year Colorectal Cancer Screening Demonstration Program (CRCSDP) in 2005 for low-income, under-insured or uninsured men and women aged 50 to 64 years in 5 participating US program sites. In this report, the authors describe the design of the CRCSDP and the overall clinical findings and screening test performance characteristics, including the positive fecal occult blood testing (FOBT) rate; the rates of polyp, adenoma, and cancer detection with FOBTs and colonoscopies; and the positive predicative value for polyps, adenomas, and cancers.


In total, 5233 individuals at average risk and increased risk were screened for colorectal cancer across all 5 sites, including 44% who underwent screening FOBT and 56% who underwent screening colonoscopy. Overall, 77% of all individuals screened were women. The FOBT positivity rate was 10%. Results from all screening or diagnostic colonoscopies indicated that 75% had negative results and required a repeat screening colonoscopy in 10 years, 16% had low-risk adenomas and required surveillance colonoscopy in 5 to 10 years, 8% had high-risk adenomas and required surveillance colonoscopy in 3 years, and 0.6% had invasive cancers.


This report documents the successes and challenges in implementing the CDC's CRCSDP and describes the clinical outcomes of this 4-year initiative, the patterns in program uptake and test choice, and the comparative test performance characteristics of FOBT versus colonoscopy. Patterns in final outcomes from the follow-up of positive screening tests were consistent with national registry data. Cancer 2013;119(15 suppl):2820–33. © 2013 American Cancer Society.


The Centers for Disease Control and Prevention (CDC) designed and established a 4-year Colorectal Cancer Screening Demonstration Program (CRCSDP) in 2005 for low-income, under-insured or uninsured men and women ages 50 to 64 years in 5 participating US program sites. This report documents the successful implementation of the CRCSDP and describes the clinical outcomes of this 4-year initiative, the patterns of program uptake and test choice, the challenges of implementing fecal occult blood testing (FOBT) screening programs, and the comparative test performance characteristics of FOBT versus colonoscopy.

Cancer is the second leading cause of death among US adults, and colorectal cancer is the second leading cause of US cancer-related deaths among cancers that affect both men and women. In 2007, 142,672 new cases and 53,219 deaths occurred from colorectal cancer.[1] Although this disease affects both sexes and individuals of all races and ethnicity, men, African Americans, and Alaska Natives have a disproportionately high incidence and mortality from colorectal cancer.[1-3] Colorectal cancer is also expensive to treat; it is estimated that the national expenditure for the medical treatment of colorectal cancer was $14 billion in 2010.[4]

The natural history of this disease is well described; most colorectal cancers develop from pre-existing, premalignant polyps, some of which will slowly transform into cancers over a period of 10 to 15 years. Through colorectal screening, polyps can be detected and removed before they transform into cancer, and cancers can be detected at an earlier stage when they are more responsive to treatment.[5]

Current guidelines recommend that adults ages 50 to 75 years at average risk for colorectal cancer should be screened with 1 or more of the following methods: FOBT every year, sigmoidoscopy every 5 years with interval FOBT every 3 years, or colonoscopy every 10 years.[6] Despite substantial scientific evidence that routine screening for colorectal cancer reduces both incidence and mortality from this disease,[7-14] screening rates have been slow to increase, and disparities in screening remain.[15-17] In 2010, 65.4% of US adults were screened according to recommended guidelines compared with 64% in 2008, 61% in 2006, 57% in 2004, and 54% in 2002.[15] Although colorectal cancer screening test use has increased over time for the overall population, racial/ethnic minorities, those without health insurance, those with lower household incomes, and those with less than a high school education had substantially lower rates of test use than comparison groups.[16, 17] Increasing poverty levels were associated with decreasing colorectal cancer screening rates. Those who had not visited a physician in the past year also were less likely to be screened.[16]

To address this disparity in screening for this high burden but largely preventable disease, the CDC designed and established a 4-year CRCSDP in 2005 for low-income, under-insured or uninsured men and women ages 50 to 64 years in the US living at or below 250% to 350% of the Federal Poverty Level.[18-21] Its overall goal was to assess the feasibility of establishing a federally funded, comprehensive colorectal cancer screening program for an underserved population. Since 1990, the CDC has administered the National Breast and Cervical Cancer Early Detection Program (NBCCEDP) primarily through state, tribal, and territorial health departments, providing breast and cervical screening services to underserved US women[22]; however, before the CRCSDP, the CDC had not established a similar program for colorectal cancer. The CDC used the CRCSDP to understand better how to structure and implement population-level colorectal cancer screening, to understand the costs of implementing a colorectal cancer screening program for underserved populations, to explore which settings and program models are most viable, and to prepare for a larger scale effort based on the outcomes of the demonstration program.

With the early success of the CRCSDP, the CDC received increased funds in 2009 to expand to a larger Colorectal Cancer Control Program, which was initiated in the fall of 2009 to coincide with the scheduled close-out of the CRCSDP. This report describes the design and implementation of the CRCSDP and the clinical outcomes of screening and follow-up tests provided by the 5 participating sites.


Five of 39 applicants were competitively selected in 2005 and awarded a total of $2.7 million per year across all 5 programs during the 4-year initiative; an accompanying article in this supplement of Cancer details annual federal funding for that program.[23] Each site differed in structure, reach, and primary screening tests and included the Maryland Department of Health and Mental Hygiene, the Missouri Department of Health and Senior Services, the Nebraska Department of Health and Human Services, Public Health-Seattle & King County, and the Stony Brook University Medical Center.[19, 20, 24-28] Because the priority population of low-income, under-insured or uninsured men and women ages 50 to 64 years represented a subset of a geographic region, some of which were small (less than a county), census data did not allow for the ability to describe the size of the priority population for the sites. Federal funds supported comprehensive program activities in all 5 sites, including clinical service delivery (screening, surveillance, and diagnostic services), patient support, data collection and tracking, program management, public education and outreach, establishment of quality services measures, maintenance of relevant partnerships, and evaluation of program process and effectiveness.

Each program was required to convene a medical advisory board to help oversee program start-up and implementation and provide guidance on clinical policy development, program eligibility criteria, patient flow, data collection, and quality assurance. Programmatic screening and diagnostic follow-up services were provided to asymptomatic individuals only. Those with gastrointestinal symptoms were referred to specialty care outside of the CRCSDP through existing provider networks, and some of the programs provided educational materials to individuals who presented with symptoms.

Although the focus of this effort was on average-risk individuals, programs could choose to offer services to individuals at increased risk because of personal or family history of adenomatous polyps or colorectal cancer. Individuals at high risk because of inflammatory bowel disease or known or suspected genetic disorders associated with colorectal cancer were referred to specialty care outside of the program. Federal funds were not used to pay for treatment for any cancers detected through this program or costs of any unintended complications associated with screening tests; these costs were secured by the screening sites themselves. Clinical complications that occurred in association with this program were rare and are described in an accompanying article.[29]

Programs were given the flexibility to select from any US Preventive Services Task Force-recommended colorectal cancer screening test[6] and were encouraged to design individual screening delivery systems that best suited their locality. At the time the CRCSDP was being established, the US Preventive Services Task Force recommended colorectal cancer screening either with annual FOBT, sigmoidoscopy, or double-contrast barium enema every 5 years or with colonoscopy every 10 years without an upper age limit.[30] Programs that selected FOBT as their primary screening test for average-risk individuals offered screening colonoscopy to individuals at increased risk from personal or family history and provided colonoscopies as diagnostic follow-up for positive FOBTs. Unless specified, FOBT cards were distributed in person at participating clinical sites.

A detailed description of the service delivery component of this program and patient recruitment are included in accompanying articles in this supplement.[24-28] Briefly, the 5 CRCSDP programs were structured as follows:

  • The Maryland program (referred to as Baltimore City throughout this article) served residents of Baltimore City and offered colonoscopy as a primary screening test.[24] The Maryland Department of Health and Mental Hygiene contracted with 5 Baltimore hospitals, where patient enrollment occurred and clinical services were provided. Patients also were recruited through the Baltimore Medical Systems, the Maryland Breast and Cervical Cancer Screening Program, and clinics affiliated with the hospitals. Treatment resources for the cancers detected in this program were provided through Medicaid, the Cigarette Restitution Fund Program, and/or charity care.
  • The Missouri program (referred to as St. Louis throughout this article) served residents of the City of St. Louis and the surrounding metropolitan area and offered screening with guaiac-based FOBT during the first 2 program years and colonoscopy during the last 2 program years. Colonoscopy also was used for high-risk individuals and in follow-up for positive FOBTs throughout the program. The Missouri Department of Health and Senior Services contracted with St. Louis ConnectCare, a federally qualified health center in St. Louis, to provide overall program coordination, recruitment, and clinical services, including colonoscopy. Some colonoscopy services also were provided by the Barnes Jewish Hospital/Siteman Cancer Center of Washington University of St. Louis and the Endoscopy Center of St. Louis. Treatment for cancers and for any unintended complications was provided to program-eligible clients at no cost through a partnership with Siteman Cancer Center and the Barnard Cancer Institute.
  • The Nebraska program served residents throughout the state and offered guaiac-based FOBT (Hemoccult Sensa II; Beckman Coulter, Inc., Jersey City, NJ) as a primary screening test, with colonoscopy primarily used for individuals with increased risk and in follow-up for positive FOBTs. Clients initially were recruited from the Nebraska Breast and Cervical Cancer Early Detection Program (Nebraska BCCEDP), and efforts were made to reach men in the families of women enrolled in the Nebraska BCCEDP. Over time, additional recruitment efforts were implemented specifically targeting men. FOBT cards were sent by mail, and the Nebraska Department of Health and Human Services used state health department staff to provide and track FOBT distribution and completion. The health department contracted services for program data entry. Program colonoscopies were provided by gastroenterologists, family practitioners (primarily in rural Nebraska), and some surgeons on a contractual fee-for-service basis. All cancer centers in the state signed letters of commitment agreeing to use charity care programs to provide treatment for any client diagnosed with cancer through the program.
  • The New York program (referred to as Suffolk County, NY throughout this article) served residents of Suffolk County on Long Island and offered colonoscopy as a primary screening test.[25, 26] The Suffolk County, NY program was administered by the Stony Brook University Medical Center in close collaboration with the Suffolk County Health Department. Clients were recruited from Suffolk County Health Centers, and all clinical care was provided by clinicians within the Stony Brook University Medical Center. This program used a full-time, onsite patient navigator and a lead clinician who worked with potential clients to provide education, determine eligibility, and describe the process to prepare for colonoscopy. Hospital financial assistance was provided to patients for treatment, and up to 95% of hospital charges related to treatment were forgiven. The university colorectal surgeons waived their professional fee for treating patients based on an agreement reached at the time of the application for the CRCSDP.
  • The Washington program (referred to as Greater Seattle throughout this article) served clients living in Clallam, Jefferson, and King Counties and offered high-sensitivity, guaiac-based FOBT as the primary screening test for average-risk individuals and colonoscopy for individuals at increased risk and for follow-up of positive FOBTs.[27] The program was administered by Public Health-Seattle & King County, and contracts were established with primary and specialty care centers to assess screening eligibility, deliver FOBT services, provide patient navigation services to those who were referred for colonoscopy, provide colonoscopies, ensure tracking and follow-up, and provide laboratory services. Clients were recruited from the Washington BCCEDP program as well as from new clinical sites that were contracted for the purposes of the CRCSDP.[27] Midway through the demonstration program, the Greater Seattle program began to offer “opt-in” colonoscopy, in which clients could select either FOBT or colonoscopy for screening after discussion with clinic staff. Treatment was paid for by Medicaid and charity care from the hospital treatment center, with first year treatment costs reaching nearly $300,000.

Each of the 5 programs used components of existing service delivery programs (eg, the NBCCEDP) or tailored program components to be integrated into existing clinical structures, including the use of a common client in-reach or out-reach system, modifying an existing clinical data collection system for use in the program, or including some members from pre-existing medical advisory boards on the CRCSDP medical advisory board.

Program Materials

During the start-up prescreening phase, which ranged from 9 months to 11 months across the 5 programs, various program materials were developed collaboratively by CDC and program staff at each site. These included program policies, a readiness checklist, a standardized set of data elements (Colorectal Cancer Clinical Data Elements [CCDEs]) with an accompanying Data User's Manual, data collection forms, a cost assessment tool, and an evaluation plan ( Data collection forms were individualized by each program site and are available on request. Programs developed awareness campaigns, and several chose to use CDC's Screen for Life: A National Colorectal Cancer Action Campaign materials (available at:

Program policies

CDC and individual site-specific program policy manuals were created for this demonstration program, as previously described.[18] Program policies addressed patient eligibility, reimbursement, reporting of complications, data and reporting, and medical advisory committees. The CDC developed service quality indicators to capture the characteristics of the population being served in the program, appropriateness of screening tests, completeness of clinical follow-up and treatment, and timeliness of clinical follow-up and treatment.

Readiness checklist

The readiness checklist was used by the CDC and the programs during start-up to monitor progress toward implementation, including planning for screening and diagnostic follow-up services, public education and outreach, data collection and tracking, patient support, partnership development and maintenance, quality assurance and professional development, and program management. Other states that were not funded under the CRCSDP used the checklist, which was made available online, to establish state-based initiatives.

Clinical Data Collection

A standardized data dictionary, referred to as the CCDEs, was developed by the CDC, the program sites, and 2 invited clinical experts and was used to describe clinical program activities. The program sites developed internal software systems to collect data on each client screened through the program. CCDE data were submitted to the CDC on a quarterly basis and were used to monitor performance, ensure service quality, develop site-specific technical assistance, and guide future policy and program development. A data contractor, Information Management Services, Inc., provided technical assistance to the CDC and the program sites to standardize, clean, validate, aggregate, and analyze the data. Tools that were developed to promote standardized reporting included a communication web site for program participants and software to validate the submission data set. Feedback reports were generated from each data submission, and each data submission was reviewed with each site to identify and correct programmatic, clinical, and data-quality issues. Efforts to measure and emphasize the importance of delivering high-quality clinical services through this program are described in an accompanying report.[31] The clinical data included in this report were collected through the CCDEs. Several data definitions were used for the CCDEs, including the following:

Screening cycle

Each CCDE record represented a “screening cycle,” which tracked a client from an initial colorectal cancer test through all follow-up tests recommended to reach a diagnosis and/or to start treatment for cancers diagnosed. The screening cycle structure helped to assess data completeness and ensure service quality. Clients who did not complete recommended follow-up were classified as “refused” when their relationship with the program was severed, including those who received follow-up outside of the program or who moved from the catchment area; “lost to follow-up” if they were nonresponsive; or “pending/unknown” when their final status was unavailable. Rescreening or surveillance tests were reported in a new CCDE record using a unique patient identifier to link clients to their subsequent rescreening or surveillance tests. Data were collected on adequacy of bowel preparation, whether the cecum was reached, complications, whether a biopsy or polypectomy was performed, the number of specimens sent to pathology, and completeness of polyp removal. A “final outcome” was computed for each cycle based on the reported diagnostic status and result.

Test indication

Test indication was recorded by the program sites for each individual who received a test within the program. Screening tests were provided for individuals at average or unknown risk and for those with a family history of colorectal cancer or adenoma. Four of the 5 programs chose to offer surveillance colonoscopies for those with a personal history of colorectal polyps or cancer.

Polyp findings

Endoscopy and pathology findings were reported for up to 15 individual polyps/colorectal lesions, including location, size, procedures for removal, completeness of removal, and histology. If more than 15 polyps were detected, then the polyps were classified according to the most severe histology. We considered a client as having advanced adenoma if they had any of the following: any adenoma that measured ≥1 cm, multiple (3 or more) adenomas of any size, or an adenoma with villous features or high-grade dysplasia.


Treatment status and date represented the date that treatment was initiated, including the date of polypectomy if no further treatment was needed, the date that the individual underwent surgical resection, or the date radiation or chemotherapy was initiated.

Data linkage

Program sites were required to link CRCSDP-diagnosed cases of high-grade dysplasia or cancers with their state central cancer registry to confirm diagnoses reported through the program and to collect registry-standardized stage data. Discrepancies between program and registry data were reported (n = 3 with registry data that reported a lower stage). Registry linkages were incomplete at program close-out for 7 of 18 linked cases because of typical delays in registry data capture. When unavailable through data linkage, stage information was provided from program records.


Unintended complications that occurred within 30 days after test provision were captured through the CCDEs and a supplemental complications reporting form and are described in an accompanying article in this journal supplement.[29]

The data reported here represent all clients who received a screening, surveillance, or rescreening test (for FOBT only) that was provided through the program. Although surveillance and rescreening recommendations were made for those who received screening colonoscopy, the program concluded before most rescreening intervals occurred.


A comprehensive evaluation of all components of the CRCSDP implementation is described in accompanying reports in this supplement. Qualitative and quantitative, multisite and multiple-method evaluation approaches were used to assess program development, overall facilitators, and challenges associated with the implementation of this program, clinical outcomes, and costs.[23, 28, 29, 31-34]

Data Analysis

The FOBT positivity rate was calculated as the percentage of FOBT tests that were positive among those individuals who completed the screening test. Polyp, adenoma, and cancer detection rates were calculated for completed FOBTs and colonoscopies as the percentage of the population who received screening colonoscopy or diagnostic colonoscopy (when FOBT was the screening test) that had polyps, adenomas, and cancers, respectively.[35] The positive predictive value for polyps, adenomas, and cancers was calculated for FOBTs as the ratio of FOBTs in which polyps, adenomas, or cancers were detected to the total number of diagnostic colonoscopies performed.[35]


In total, 5233 individuals received colorectal cancer screening tests across all 5 sites, including those at average and increased risk. Annually, 130 new clients (in program year 1), 1376 new clients (in program year 2), 2048 new clients (in program year 3) and 1679 new clients (in program year 4) were screened. Year 1 screening numbers reflect the fact that most of the year was devoted to prescreening start-up. After an increase from year 1 to year 3, screening numbers declined because of scheduled program close-out. The proportion of total screening tests received in program years 1, 2, 3, and 4 were 47%, 48%, 48%, and 73%, respectively, for colonoscopy; 53%, 52%, 52%, and 26%, respectively, for FOBT; and <1% in all 4 years for other tests. Table 1{TBL 1} lists the characteristics of the participants who received screening and diagnostic services through the CRCSDP. Of the 5233 individuals screened, 2294 (44%) received screening FOBT, and 2937 (56%) received screening colonoscopy.

Table 1. Characteristics of Participants Receiving Screening and Follow-Up Services in the Colorectal Cancer Screening Demonstration Program by Program Site and Overall, 2005–2009
Participant CharacteristicsPercentage of Participants (%)
Baltimore CitySt. LouisNebraskaSuffolk County, NYGreater SeattleTotal
  1. Abbreviations: CRC, colorectal cancer; FOBT, fecal occult blood test.

All participants696 (13.3)426 (8.1)2025 (38.7)800 (15.3)1286 (24.6)5233 (100)
Primary screening test
FOBT0 (0)67 (15.7)1319 (65.1)0 (0)908 (70.6)2294 (43.8)
Colonoscopy696 (100)359 (84.3)706 (34.9)800 (100)376 (29.2)2937 (56.1)
Other/sigmoidoscopy0 (0)0 (0)0 (0)0 (0)2 (0.2)2 (0)
Men190 (27.3)125 (29.3)248 (12.2)281 (35.1)367 (28.5)1211 (23.1)
Women506 (72.7)301 (70.7)1777 (87.8)519 (64.9)919 (71.5)4022 (76.9)
Age, y
<5021 (3)35 (8.2)10 (0.5)23 (2.9)38 (3)127 (2.4)
50–59486 (69.8304 (71.41437 (71)585 (73.1)899 (69.9)3711 (70.9)
60–64177 (25.483 (19.5456 (22.5)135 (16.9)272 (21.2)1123 (21.5)
≥6512 (1.74 (0.9122 (6)57 (7.1)77 (6)272 (5.2)
Hispanic25 (3.6)11 (2.6)421 (20.8)362 (45.3)239 (18.6)1058 (20.2)
Non-Hispanic671 (96.4)415 (97.4)1604 (79.2)438 (54.8)1047 (81.4)4175 (79.8)
White98 (14.1)170 (39.9)1423 (70.3)235 (29.4)504 (39.2)2430 (46.4)
Black548 (78.7)237 (55.6)114 (5.6)144 (18)169 (13.1)1212 (23.2)
Asian/Pacific Islander7 (1)3 (0.7)20 (149 (6.1247 (19.2326 (6.2
American Indian/Alaska Native1 (0.1)1 (0.2)16 (0.85 (0.663 (4.986 (1.6
Multiple/unknown17 (2.4)4 (0.9)31 (1.5)5 (0.6)64 (5)121 (2.3)
Risk for CRC
Average/unknown risk551 (79.2)378 (88.7)1541 (76.1)718 (89.8)1160 (90.2)4348 (83.1)
Increased risk145 (20.8)48 (11.3)484 (23.9)82 (10.3)126 (9.8)885 (16.9)
Family history133 (19.1)42 (9.9)452 (22.3)82 (10.3)102 (7.9)811 (15.5)
Personal history12 (1.7)6 (1.4)32 (1.6)0 (0)24 (1.9)74 (1.4)
Screening history
Yes, any test149 (21.4)91 (21.4)801 (39.6)297 (37.1)236 (18.4)1574 (30.1)
FOBT86 (12.4)61 (14.3)604 (29.8)251 (31.4)124 (9.6)1126 (21.5)
Sigmoidoscopy11 (1.6)9 (2.1)58 (2.9)34 (4.3)28 (2.2)140 (2.7)
Colonoscopy60 (8.6)37 (8.7)304 (15)39 (4.9)106 (8.2)546 (10.4)
None/unknown547 (78.6)335 (78.6)1224 (60.4)503 (62.9)1050 (81.6)3659 (69.9)

Overall, 77% of all individuals screened were women, and this proportion ranged from 65% (Suffolk County, NY) to 88% (Nebraska). The majority of participants (92%) were ages 50 to 64 years, and 71% were ages 50 to 59 years. Across the 5 programs, 46% of those screened were non-Hispanic white, although race and ethnicity varied by program site. Those reporting African American or black race represented 79% of the population screened in Baltimore City compared with 5.6% in Nebraska, Asians/Pacific Islanders represented 19% of those screened in Greater Seattle compared with <1% in St. Louis, and American Indians/Alaska Natives represented 5% of the total population screened in Greater Seattle compared with <1% in all other program sites. Individuals of Hispanic ethnicity represented 20% of the total number screened, ranging from 2.6% (St. Louis) to 45% (Suffolk County, NY).

Consistent with the intent of the program, the majority (83%) of those who received screening services were at average risk or unknown risk for colorectal cancer. The proportion of individuals receiving tests who were at increased risk for colorectal cancer because of a family or personal history varied by program, reflecting the program site-specific policy regarding this eligibility criterion. Across all programs, most individuals (70%) who received services were previously unscreened. Of those who had been previously screened, 22% had been screened with FOBT, 10% had been screened with colonoscopy, and 3% had been screened with sigmoidoscopy.

Figure 1 illustrates the outcomes of screening test distribution, completion, and diagnostic follow-up for the 3 program sites that used FOBT for primary screening (Greater Seattle, Nebraska, and St. Louis). Of the 4366 FOBT kits that were distributed across all programs, 47% or 2072 kits were not returned (range, 37%-53% of FOBT kits). Of the 2294 individuals who returned their kits, 42 did not complete the FOBT screening cycle because it was determined that they were at increased risk for colorectal cancer and, thus, needed a colonoscopy instead of an FOBT. Of the remaining 2252 individuals at average risk who returned their FOBT cards, 2027 or 90% of tests were negative, for an overall FOBT positivity rate of 10%. Those with a negative FOBT were recommended for rescreening with FOBT in 1 year. Rescreening rates of 13% using FOBT and 16% using FOBT or colonoscopy were based on 1516 clients whose initial FOBT occurred within a 12-month to 30-month interval across the 5 program sites, providing enough time to be rescreened through the program. Of those who were rescreened with FOBT, 180 (94%) had a negative result, 9 received follow-up with colonoscopy, and 2 individuals had advanced adenoma identified.

Figure 1.

Testing cascade and clinical outcomes are illustrated for participants in the Colorectal Cancer Screening Demonstration Program (CRCSDP) who received primary screening with the fecal occult blood test (FOBT). A single asterisk indicates that FOBT was provided before it was determined that patients were at increased risk. Among these, 10 participants had positive FOBT results and were recalled for diagnostic colonoscopy, and 32 had negative FOBT results, including 4 who were recalled for screening colonoscopy. A double asterisk indicates that the rescreened participants included clients who were rescreened 12 to 30 months after the initial negative FOBT. The percentage rescreened is based on 1516 clients whose annual rescreen date occurred before the conclusion of CRCSDP screening. DCBE indicates double-contrast barium enema; Dx, diagnostic.

Among the 225 individuals who had a positive FOBT, 82% underwent a follow-up colonoscopy, <1% underwent follow-up with other tests, and 16% were lost to follow-up with no additional records after their initial positive FOBT as of program close-out. The results of the diagnostic colonoscopies included 144 normal examinations, 9 with advanced adenomas, and 4 with invasive cancers.

Table 2 lists the outcomes of the 3 FOBT screening programs combined according to sex, age, race/ethnicity, and screening history. The proportion with negative tests was the same for both men and women (90% each), although the results of diagnostic follow-up colonoscopy varied by sex. Seventy percent of women had a negative colonoscopy after a positive FOBT versus 33% of men. Men had a higher rate of adenomas (31% vs 16% in women) and invasive cancers (5% vs 1% in women) than women, although the small numbers of any individual outcomes limited the reliability of these proportions. The prevalence of advanced adenoma increased with increasing age.

Table 2. Clinical Outcomes of Average-Risk Colorectal Cancer Screening Demonstration Program Participants Receiving Primary Screening with Fecal Occult Blood Testing in St. Louis, Nebraska, and Greater Seattle, 2005–2009
Participant CharacteristicsNo. of Participants (%)
TotalScreen ResultOutcome of Diagnostic Follow-Up Colonoscopya
Negative FOBTPositive FOBTNormalAll AdenomaAdvanced AdenomaCancerLost/Refused/Pending/Unknown
  1. Abbreviations: FOBT, fecal occult blood test;

  2. a

    This analysis included 4 individuals who reached a final diagnosis through tests other than colonoscopy.

All2252 (100)2027 (90)225 (10)144 (64)41 (18.2)9 (4)4 (1.8)36 (16)
Men393 (17.5)354 (90.1)39 (9.9)13 (33.3)12 (30.8)2 (5.1)2 (5.1)12 (30.8)
Women1859 (82.5)1673 (90)186 (10)131 (70.4)29 (15.6)7 (3.8)2 (1.1)24 (12.9)
Age, y
<5022 (1)19 (86.4)3 (13.6)0 (0)3 (100)3 (100)0 (0)0 (0)
50–591584 (70.3)1435 (90.6)149 (9.4)102 (68.5)22 (14.8)3 (2)2 (1.3)23 (15.4)
60–64496 (22)437 (88.1)59 (11.9)37 (62.7)13 (22)2 (3.4)2 (3.4)7 (11.9)
≥65150 (6.7)136 (90.7)14 (9.3)5 (35.7)3 (21.4)1 (7.1)0 (0)6 (42.9)
Hispanic548 (24.3)510 (93.1)38 (6.9)25 (65.8)6 (15.8)2 (5.3)0 (0)7 (18.4)
White1163 (51.6)1045 (89.9)118 (10.1)89 (75.4)19 (16.1)7 (5.9)2 (1.7)8 (6.8)
Black210 (9.3)186 (88.6)24 (11.4)11 (45.8)3 (12.5)0 (0)2 (8.3)8 (33.3)
Asian/ Pacific Islander198 (8.8)167 (84.3)31 (15.7)12 (38.7)10 (32.3)0 (0)0 (0)9 (29)
American Indian/Alaska Native64 (2.8)59 (92.2)5 (7.8)3 (60)1 (20)0 (0)0 (0)1 (20)
Multiracial/unknown race69 (3.1)60 (87)9 (13)4 (44.4)2 (22.2)0 (0)0 (0)3 (33.3)
Previous FOBT within 2 y
Yes248 (11)226 (91.1)22 (8.9)17 (77.3)3 (13.6)1 (4.5)0 (0)2 (9.1)
No/unknown2004 (89)1801 (89.9)203 (10.1)127 (62.6)38 (18.7)8 (3.9)4 (2)34 (16.7)

Table 3 provides the results of screening colonoscopies performed in the CRCSDP from all 5 program sites. In total, 2868 individuals received screening colonoscopy, including 2095 at average risk and 773 at increased risk because of family history. Among those who were screened with colonoscopy, 207 were diagnosed with advanced adenomas, 14 were diagnosed with invasive cancers, and 43 (1.5%) were lost to follow-up before a final diagnosis was reached. The total numbers of neoplastic lesions were higher than those detected by FOBT. Subgroup patterns were similar to the patterns observed in the FOBT screening cohort, with a higher prevalence of adenomas and cancers among men compared with women and the prevalence of adenoma increasing with increasing age. Adenoma rates were higher among individuals who had a family history compared with those at average risk. In total, 69 individuals received surveillance colonoscopy as their first test in the CRCSDP because of a history of colorectal polyps or cancer. A higher prevalence of adenomas was observed among those who underwent surveillance colonoscopy compared with screening colonoscopy, and no cancers were detected in the surveillance group.

Table 3. Clinical Outcomes of Colorectal Cancer Screening Demonstration Program Participants Screened With Colonoscopy in all Program Sites, 2005–2009a
Participant CharacteristicTotal No. of ParticipantsOutcome: No. of Participants (%)
NormalAll AdenomaAdvanced AdenomaInvasive CancerLost/Refused/Pending/Unknown
  1. a

    These did not include individuals who received colonoscopy to follow a positive fecal occult blood test.

Screening cohort
All screening colonoscopies28682100 (73.2)711 (24.8)207 (7.2)14 (0.5)43 (1.5)
Average or unknown risk20951547 (73.8)502 (24)146 (7)10 (0.5)36 (1.7)
Family history of colorectal cancer773553 (71.5)209 (27)61 (7.9)4 (0.5)7 (0.9)
Men794519 (65.4)257 (32.4)93 (11.7)6 (0.8)12 (1.5)
Women20741581 (76.2)454 (21.9)114 (5.5)8 (0.4)31 (1.5)
Age, y      
<509977 (77.8)21 (21.2)5 (5.1)0 (0)1 (1)
50–5920631530 (74.2)492 (23.8)136 (6.6)11 (0.5)30 (1.5)
60–64590413 (70)163 (27.6)55 (9.3)3 (0.5)11 (1.9)
≥6511680 (69)35 (30.2)11 (9.5)0 (0)1 (0.9)
Hispanic503374 (74.4)122 (24.3)29 (5.8)2 (0.4)5 (1)
Non-Hispanic23651726 (73)589 (24.9)178 (7.5)12 (0.5)38 (1.6)
White1186845 (71.2)323 (27.2)83 (7)7 (0.6)11 (0.9)
Black986753 (76.4)206 (20.9)81 (8.2)3 (0.3)24 (2.4)
Asian/ Pacific Islander12383 (67.5)38 (30.9)10 (8.1)1 (0.8)1 (0.8)
American Indian/Alaska Native2115 (71.4)5 (23.8)1 (4.8)0 (0)1 (4.8)
Multiracial/unknown race4930 (61.2)17 (34.7)3 (6.1)1 (2)1 (2)
Surveillance cohort6946 (66.7)23 (33.3)6 (8.7)0 (0)0 (0)

By program site (data not shown), 52 advanced adenomas and 3 cancers were detected in the Baltimore City program (primary colonoscopy screening); 33 advanced adenomas and 4 cancers were detected in the St. Louis program (mixed test options); 43 advanced adenomas and 5 cancers were detected in the Nebraska program (primary FOBT screening); 51 advanced adenomas and 5 cancers were detected in the Suffolk County, NY program (primary colonoscopy screening); and 45 advanced adenomas and 1 cancer were detected in the Greater Seattle program (mixed test options). Among the 18 cancers diagnosed in the CRCSDP, 4 were stage 0, 3 were stage I, 3 were stage IIA, 2 were stage III, 1 was stage IV, and stage data were missing for 5 of the cancers. Treatment was initiated for all patients.

Table 4 lists provider, practice site, and procedure characteristics for colonoscopies performed for the screening cohort (primary screening colonoscopies in those with average and increased risk and colonoscopies to follow-up a positive screening FOBT). Provider specialty varied by program site. Gastroenterologists performed the majority of colonoscopies across all programs, ranging from 100% (Suffolk County, NY) to 36% (Nebraska). All programs except for Suffolk County, NY used a mix of endoscopists, with family practitioners, general surgeons, colorectal surgeons, and internists following gastroenterologists as most the frequent specialists performing colonoscopy. Clinical practice site also varied by program. Procedures were most frequently performed in hospitals and ambulatory surgery/endoscopy centers, with the proportion performed in hospitals ranging from 0% (St. Louis, Mo) to 100% (Suffolk County, NY) and the proportion performed in ambulatory surgery/endoscopy centers ranging from 0% (Suffolk County, NY) to 100% (St. Louis). On average, across all programs, 44% of procedures led to a biopsy, with some variability across programs, and snare polypectomy and cold biopsy were the most common methods of tissue removal.

Table 4. Characteristics of Providers, Provider Sites, and Procedures in the Colonoscopy Screening Cohort: Colorectal Cancer Screening Demonstration Program, 2005–2009a
CharacteristicsNo. (%)
Baltimore CitySt. LouisNebraskaSuffolk County, NYGreater SeattleAll
  1. Abbreviations: FOBT, fecal occult blood test.

  2. a

    This includes screening colonoscopies performed in individuals at average risk for colorectal cancer, those with a family history of colorectal cancer, and colonoscopies performed to follow a positive FOBT (N = 3145).

All colonoscopies684 (100)364 (100)792 (100)800 (100)505 (100)3145 (100)
Colonoscopy screening cohort
Follow-up of positive FOBT0 (0)5 (1.4)76 (9.6)0 (0)129 (25.5)210 (6.7)
Screening colonoscopy, average risk551 (80.6)319 (87.6)261 (33)718 (89.8)293 (58)2142 (68.1)
Screening colonoscopy, family history133 (19.4)40 (11)455 (57.4)82 (10.3)83 (16.4)793 (25.2)
Provider specialty
General practitioner0 (0)0 (0)9 (1.1)0 (0)0 (0)9 (0.3)
Internist0 (0)0 (0)60 (7.6)0 (0)0 (0)60 (1.9)
Family practitioner3 (0.4)0 (0)205 (25.9)0 (0)0 (0)208 (6.6)
Gastroenterologist549 (80.3)279 (76.6)289 (36.5)800 (100)425 (84.2)2342 (74.5)
General surgeon44 (6.4)8 (2.2)167 (21.1)0 (0)0 (0)219 (7)
Colorectal surgeon88 (12.9)77 (21.2)61 (7.7)0 (0)44 (8.7)270 (8.6)
Nurse practitioner0 (0)0 (0)0 (0)0 (0)0 (0)0 (0)
Physician assistant0 (0)0 (0)1 (0.1)0 (0)0 (0)1 (0)
Unknown0 (0)0 (0)0 (0)0 (0)36 (7.1)36 (1.1)
Clinical practice site
Physician's office0 (0)0 (0)2 (0.3)0 (0)0 (0)2 (0.1)
Ambulatory endoscopy/surgery center185 (27)364 (100)401 (50.6)0 (0)386 (76.4)1336 (42.5)
Hospital496 (72.5)0 (0)389 (49.1)800 (100)99 (19.6)1784 (56.7)
Health clinic2 (0.3)0 (0)0 (0)0 (0)1 (0.2)3 (0.1)
Unknown1 (0.1)0 (0)0 (0)0 (0)19 (3.8)20 (0.6)
Biopsy/polypectomy performed?
Yes251 (36.7)160 (44)330 (41.7)380 (47.5)276 (54.7)1397 (44.4)
No/unknown433 (63.3)204 (56)462 (58.3)420 (52.5)229 (45.3)1748 (55.6)
Procedure used in removal/biopsy
Snare polypectomy92 (13.5)72 (19.8)151 (19.1)217 (27.1)85 (16.8)617 (19.6)
Hot biopsy forceps/cautery91 (13.3)72 (19.8)48 (6.1)77 (9.6)30 (5.9)318 (10.1)
Cold biopsy106 (15.5)83 (22.8)134 (16.9)121 (15.1)197 (39)641 (20.4)
Ablation0 (0)0 (0)2 (0.3)0 (0)2 (0.4)4 (0.1)

Table 5 provides the characteristics of the lesions identified on colonoscopies performed for the screening cohort. Among the 3145 colonoscopies, 2001 (64% of the total cohort) were normal or had no neoplastic findings or hyperplastic polyps; and, in another 361 colonoscopies (11.5%), hyperplastic polyps were identified. In 654 individuals (21%), the most advanced lesion was a tubular adenoma, and the majority of those were small polyps that measured <1 cm. In a total of 243 individuals, colonoscopies revealed multiple polyps (range, 3-10 polyps), tubular adenomas ≥1 cm, adenomas with villous features, serrated polyps (2 of whom had multiple polyps or tubulovillous tissue identified), or high-grade dysplasia; and 18 individuals (0.6% of the total cohort) had invasive cancer. On the basis of the current surveillance guidelines, 2362 individuals (75%) had non-neoplastic findings on screening colonoscopy requiring a rescreen with colonoscopy in 10 years, 500 individuals (16%) had low-risk adenomas requiring surveillance colonoscopy in 5 to 10 years, and 243 (8%) had high-risk adenomas requiring surveillance colonoscopy in 3 years.[36] By location, approximately 33% of the colonoscopies in which only hyperplastic polyps were identified revealed polyps in the rectum only; and 45% of all colonoscopies that revealed a histology of adenoma or higher grade neoplasia identified adenomas in the distal colon only, 37% identified adenomas in the proximal colon only, and 16% identified adenomas in both the distal colon and the proximal colon.

Table 5. Histologic Characteristics in the Colonoscopy Screening Cohort: Colorectal Cancer Screening Demonstration Program, 2005–2009 (n = 3145)
HistologyNo. of Participants (%)
  1. Abbreviations: NOS, not otherwise specified.

  2. a

    These excluded hyperplastic polyps.

  3. b

    Counts of tubular adenoma included a reported histology of adenoma NOS.

  4. c

    Of the clients who had serrated adenomas, 1 had >10 adenomas, and 1 had 1 or 2 tubulovillous >1 cm.

  5. d

    Lesions included histology of adenoma or worse.

Normal2001 (63.6)
No polyps1780 (89)
Normal/nonpolyp181 (9)
Nonadenomatous polypsa40 (2)
Hyperplastic361 (11.5)
Other locations230 (63.7)
Rectal only131 (36.3)
Tubular/NOS adenomab654 (20.8)
Unknown size, 1–2 lesions22 (3.4)
Unknown size, 3–10 lesions1 (0.2)
<1 cm, 1–2 lesions478 (73.1)
<1 cm, 3–10 lesions61 (9.3)
≥1 cm, 1–2 lesions89 (13.6)
≥1 cm, 3–10 lesions3 (0.5)
Tubulovillous adenoma45 (1.4)
Unknown size, 1–2 lesions2 (4.4)
<1 cm, 1–2 lesions19 (42.2)
<1 cm, 3–10 lesions1 (2.2)
≥1 cm, 1–2 lesions22 (48.9)
≥1 cm, 3–10 lesions1 (2.2)
Villous adenoma7 (0.2)
<1 cm7 (100)
Serrated adenomac25 (0.8)
Unknown size2 (8)
<1 cm21 (84)
≥1 cm2 (8)
High-grade dysplasia12 (0.4)
Unknown size2 (16.7)
High-grade dysplasia <1 cm4 (33.)3
High-grade dysplasia ≥1 cm6 (50)
Cancer18 (0.6)
Other cancer6 (33.3)
Adenocarcinoma invasive12 (66.7)
Unclassified/incomplete data22 (0.7)
Location of lesionsd 
Distal only340 (44.7)
Proximal only280 (36.8)
Distal and proximal124 (16.3)
Unknown17 (2.2)

Table 6 lists the screening test indicators and test performance measures for FOBT and colonoscopy that were used for the CRCSDP screening cohort. Polyp, adenoma, and cancer detection rates per 1000 individuals tested with FOBT were 29.3%, 18.5%, and 1.8%, respectively, compared with 355.5%, 243.8%, and 4.9%, respectively, for screening colonoscopy in individuals at average risk. Colonoscopy polyp, adenoma, and cancer detection rates were higher in those at increased risk for colorectal cancer. The overall FOBT positivity rate was 10%, and the positive predictive value of FOBT for polyps, adenomas, and cancers was 34.1%, 21.1%, and 2.2%, respectively.

Table 6. Screening Test Indicators and Test Performance Measures for Tests Used in the Screening Cohort: Colorectal Cancer Screening Demonstration Program, 2005 to 2009a
 No. of Participants  Rate/Percentage (No.)
Primary TestScreenedbRisk for CRCPolypAdenomaCancer
  1. Abbreviations: CRC, colorectal cancer; FOBT, fecal occult blood test; PPV, positive predictive value.

  2. a

    These included screening FOBTs and screening colonoscopies performed in individuals at average risk for CRC, among those with a family history of CRC, and colonoscopies performed to follow a positive FOBT.

  3. b

    These excluded participants who did not complete testing: FOBT (n = 36), screening colonoscopy for those with average risk (n = 36), and screening colonoscopy for those with increased risk (n = 7).

  4. c

    The increased CRC risk was because of a family history of CRC.

  5. d

    The PPV was calculated only for those who underwent diagnostic colonoscopy (n = 185).

Detection rate per 1000 screening tests
Screening colonoscopy2825Total370.6 (1047)251.7 (711)5 (14)
 2059Average or unknown risk355.5 (732)243.8 (502)4.9 (10)
 766Increased riskc411.2 (315)272.8 (209)5.2 (4)
FOBT2216Average or unknown risk29.3 (65)18.5 (41)1.8 (4)
Positive FOBT rate and PPV: % (no.)d10 (225) 34.1 (63)21.1 (39)2.2 (4)


In this report, we document the design and successful implementation of the CDC's CRCSDP and describe the clinical outcomes of the approximately 5200 low-income, uninsured individuals who were screened through this 4-year initiative. We describe patterns in program uptake and test choice, challenges in the implementation of FOBT screening programs, and comparative test performance characteristics of FOBT versus colonoscopy, all of which are important issues to help guide future efforts in this area. The CDC used the overall findings, observations, and framework from this effort to design and initiate the larger, population-based Colorectal Cancer Control Program now in place, which reaches all men and women ages ≥50 years in 25 participating states and 4 tribes and tribal organizations ( .

The CRCSDP was successfully put into place in an environment in which the science related to colorectal cancer screening continues to rapidly evolve and during which screening recommendations were updated, because the importance of addressing this preventable cancer was so great. However, although screening numbers increased during the CRCSDP program period, the overall numbers of individuals reached were modest, reflecting the level of funding; the funding of nonclinical services; the small catchment area of most of the participating sites; the focus of the initiative on a specific, critically important, but relatively small and traditionally hard to reach population[37]; and the focus of the program on reimbursement for service delivery.

The numbers of organized screening programs nationally and internationally are growing, and data are accumulating that define standards for expected rates of particular clinical outcomes and test performance characteristics.[25, 31, 35, 38-43] In addition, efforts are being undertaken to standardize reporting for colorectal cancer screening programs in an effort to place appropriate importance on measuring the quality of these programs and to facilitate comparisons across programs.[35, 44, 45] The histologic findings in our cohort, including the proportion with hyperplastic polyps, adenomas, and invasive cancers, and the screening test performance characteristics were similar to outcomes from other programs that have published their data, with slightly lower rates of invasive cancer in our cohort compared with other published screening studies, primarily of men.[36, 39] Few if any of the other existing international programs target an underserved population. Most screening programs in other countries are true population-based efforts that have been established in countries with national health systems and universal coverage that reach individuals regardless of socioeconomic status, which limits comparability of the current population with other programs. Our cohort was too small and the number of years of screening too few to detect whether the low-income, largely previously unscreened population receiving services in the CRCSDP had a higher burden of disease than a general screening population.

Although the CRCSDP tried to reach all low-income, under-insured and uninsured men and women ages 50 to 64 years living in the 5 participating catchment areas, uptake was higher in certain populations. Of those receiving services, the majority were women. This was most likely because of the strong association between this program and the CDC's long-standing NBCCEDP, which serves only women, and is consistent with the observation—well documented in the peer-reviewed literature—that women usually seek preventive services at higher rates than men.[46-49] Just under half of all individuals served in the program were white, and the remaining individuals were a mix of other demographic groups. Wide demographic variability was observed across programs, reflecting the specific population mix and targeted outreach efforts in each participating site.[28] The program was successful in reaching those who had never been screened, which is critical, because most of the impact from colorectal cancer screening is accrued from the first screen.[50] Among those who had previously been screened, most had been screened using FOBT, consistent with screening trends from the years leading up to this initiative.[16, 17]

Men participated less in FOBT screening than women, had similar test positivity rates as women, but had higher adenoma and cancer detection rates with FOBT and colonoscopy than women. Even with the low numbers of outcomes, the pattern suggests a need for greater education and outreach efforts for men, for whom colorectal cancer incidence and mortality rates are higher than for women.[3] On the basis of the early program data that demonstrated this screening uptake disparity by sex, several CRCSDP programs increased targeted outreach to men.[28]

The influence of patient and provider preferences was evident in this effort. Midway through the 4-year program, several program sites changed their selected primary screening test. Among the 3 FOBT-based programs, 1 (St. Louis) switched to primary screening with colonoscopy in response to perceived patient preferences, and the Greater Seattle program initiated an opt-in approach, both to allow clients to choose their screening test (FOBT or colonoscopy) and because agency and provider capacity had grown enough to offer this option. Screening in Greater Seattle increased after the opt-in approach began, because more clinic sites decided to participate. Over the course of the CRCSDP, the proportion of screening colonoscopies performed overall increased compared with FOBT, but this was in part because of the programs' need to choose a 1-step screening test in the final year of scheduled close-out, so that colonoscopies for follow-up would not be generated from positive FOBTs after the program ended. These programmatic changes made by the sites reflected perceived patient and provider preferences and also may reflect a growing preference for screening colonoscopy.[27, 51]

We described 2 important programmatic challenges faced by programs that selected FOBT as a primary screening test, including low FOBT card return rates and low rescreening rates. On average, across the 3 FOBT programs, the FOBT card return rate was 53%, which was somewhat higher than reported by other international programs,[41, 42] but still was low. In addition, rescreening rates across the FOBT programs were low (range, 13%-16%), although this was in large part because the program underwent scheduled close-out before most clients were due for a rescreening. Return rates and rescreening rates are critical components of an FOBT program and require attention and adequate resource allocation for any jurisdiction that plans to use stool tests as a primary screening test, because the test can only be effective if the test cycle is completed and the test is repeated on an annual or biannual basis.[31] Future programs that choose FOBT may benefit from identifying reliable and systemized measures to ensure that the FOBT screening and rescreening cycle is completed.[52]

The FOBT positivity rate across all individuals screened in this program was 10%, which was almost 2 times higher than the rates reported in clinical trials and in other community-based and international programs.[7, 8, 11, 12, 35] This high positivity rate, as well as program-to-program variation, is discussed in another accompanying article in this supplement.[31] Specific guidance was not given in the CRCSDP on which guaiac-based stool tests the programs should use; however, as the science has evolved in this area, participants in the CDC's newer Colorectal Cancer Control Program are advised to use only high-sensitivity stool tests ( , which will increase test performance and may reduce some of the variation in FOBT positivity rates in the future.

Although the number of surveillance colonoscopies was low, the average rate of adenomas observed in this population was higher than in the screening population. Screening colonoscopy revealed higher adenoma rates in men than in women and increasing adenoma rates with increasing age, consistent with the pattern of invasive colorectal cancer by demographic subgroup observed in national registry data.[1-3]

We observed geographic differences in the types of providers performing colonoscopy, with more family practitioners and colorectal surgeons performing colonoscopy in Nebraska than in the other CRCSDP sites. This follows the pattern observed in the Survey of Endoscopic Capacity (SECAP) state-based studies ( , which demonstrated that a mix of provider types were performing colonoscopy in rural areas compared with urban centers (unpublished data). The variability by procedure site observed in the CRCSDP data reflect the planned referral patterns in each of these programs, in which, for instance, the Suffolk County program only performed colonoscopies within its participating hospital system, and the St. Louis program primarily performed colonoscopies in its participating and dedicated ambulatory surgery center.

Although there are limitations to comparing FOBT (an annual test) with colonoscopy (a test performed once every 10 years if no significant findings are identified on the initial colonoscopy) in this 4-year cohort, some important differences emerged. The lost-to-follow-up rate was lower with colonoscopy than with FOBT, which is an important consideration for a screening program. The polyp, adenoma, and cancer detection rates were much higher for colonoscopy than for FOBT, as would be expected. The observed comparative polyp, adenoma, and detection rates between these 2 tests might be factored into a given program's test selection along with other important considerations, including comparative costs, complication rates, patient and provider preferences, and capacity.[53, 54]

The CDC's CRCSDP provided critical lessons that paved the way for follow-up efforts in the United States, including CDC's population-based colorectal cancer control initiative. The continued challenges with programs now in place will be to keep a focus on the most vulnerable populations while also reaching all individuals aged ≥50 years and to ensure that current and future programs align with the Affordable Care Act.[55] Thus, as science and public health service delivery both continue to evolve, it is critical that attention remain on screening for this very common but largely preventable disease.


The Colorectal Cancer Screening Demonstration Program evaluated in this supplement was funded by the Centers for Disease Control and Prevention Funding Opportunity Number RFA AA030.


The authors made no disclosures.