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.
Moving forward: Using the experience of the CDCs' Colorectal Cancer Screening Demonstration Program to guide future colorectal cancer programming efforts
Version of Record online: 18 JUL 2013
© 2013 American Cancer Society
Special Issue: Comprehensive Evaluation of the Centers for Disease Control and Prevention's Colorectal Cancer Screening Demonstration Program, Supplement to Cancer
Volume 119, Issue Supplement S15, pages 2940–2946, 1 August 2013
How to Cite
Seeff, L. C., DeGroff, A., Joseph, D. A., Royalty, J., Tangka, F. K. L., Nadel, M. R. and Plescia, M. (2013), Moving forward: Using the experience of the CDCs' Colorectal Cancer Screening Demonstration Program to guide future colorectal cancer programming efforts. Cancer, 119: 2940–2946. doi: 10.1002/cncr.28155
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.
- Issue online: 18 JUL 2013
- Version of Record online: 18 JUL 2013
- Manuscript Accepted: 7 NOV 2012
- Manuscript Revised: 6 NOV 2012
- Manuscript Received: 12 JUL 2012
- colorectal cancer screening;
The Centers for Disease Control and Prevention (CDC) established and supported a 4-year Colorectal Cancer Screening Demonstration Program (CRCSDP) from 2005 to 2009 for low-income, under- or uninsured men and women aged 50-64 at 5 sites in the United States.
A multiple methods evaluation was conducted including 1) a longitudinal, comparative case study of program implementation, 2) the collection and analysis of client-level screening and diagnostic services outcome data, and 3) the collection and analysis of program- and patient-level cost data.
Several themes emerged from the results reported in the series of articles in this Supplement. These included the benefit of building on an existing infrastructure, strengths and weakness of both the 2 most frequently used screening tests (colonoscopy and fecal occult blood tests), variability in costs of maintaining this screening program, and the importance of measuring the quality of screening tests. Population-level evaluation questions could not be answered because of the small size of the participating population and the limited time frame of the evaluation. The comprehensive evaluation of the program determined overall feasibility of this effort.
Critical lessons learned through the implementation and evaluation of the CDC's CRCSDP led to the development of a larger population-based program, the CDC's Colorectal Cancer Control Program (CRCCP). Cancer 2013;119(15 suppl):2940–6. © 2013 American Cancer Society.
From 2005 to 2009, the Centers for Disease Control and Prevention (CDC) established and supported a 4-year Colorectal Cancer Screening Demonstration Program (CRCSDP) for low-income, under- or uninsured men and women 50-64 years of age living in the United States at or below 250%-350% of the federal poverty level.[1-13] Five sites received funding for this initiative: 1) the Maryland Department of Health and Mental Hygiene (referred to as the Baltimore city site), 2) the Missouri Department of Health and Senior Services (referred to as the St. Louis site), 3) the Nebraska Department of Health and Human Services (referred to as the Nebraska site), 4) Public Health—Seattle & King County (referred to as the greater Seattle site), and 5) the Stony Brook University Medical Center (referred to as the Suffolk County, New York, site). Throughout the demonstration period, the CDC conducted a multiple methods evaluation based on the CDC's “Framework for Program Evaluation in Public Health,” including a longitudinal, comparative case study of program implementation, analysis of client-level screening and diagnostic services outcome data, and analysis of program- and patient-level cost data. In this final report of the Cancer Supplement, we highlight findings from each of the 3 evaluation components described in accompanying articles and show through these findings the feasibility of establishing and maintaining a federally funded colorectal cancer screening program for an underserved population. In addition, we describe the development of the CDC's Colorectal Cancer Control Program (CRCCP), which was informed by lessons learned from the CRCSDP. As we parlayed the lessons from the CRCSDP—an organized, relatively small-scale program—into broader organized screening efforts across the United States, attention was paid to successes and challenges encountered during the demonstration program in 3 broad areas: implementation, clinical service delivery, and cost.
Summary of CRCSDP Findings
Program Implementation Challenges and Successes
Preexisting infrastructure, partnerships, patient recruitment, use of protocols
Methods for the qualitative case-study evaluation, the results of which are highlighted here, are fully described in several articles in this Supplement.[8, 12, 13] Glover-Kudon et al described 3 stages of CRCSDP program development, early, mid- and late implementation, each with unique implementation successes and challenges and each requiring tailored technical assistance. During early implementation, preexisting infrastructure was identified as the most important facilitator of successful program start-up. All sites were able to capitalize on other colorectal cancer screening initiatives and research and some on the preexisting infrastructure of 2 well-established CDC-funded programs, the National Breast and Cervical Cancer Early Detection Program (NBCCEDP) and the Well-Integrated Screening and Evaluation for Women Across the Nation program, which support screening for breast and cervical cancer and cardiovascular health, respectively.[16, 19] These preexisting programs provided networks of clinical providers, data management and billing systems, medical advisory boards, program policies and procedures, patient referral systems, public education materials, and program staff with valuable experience in designing, managing, and implementing public health screening programs. As 1 example, the greater Seattle site built its CDC-funded program on the existing infrastructure of its NBCCEDP program and was then able to identify state resources to scale its program to the entire state. The single CRCSDP site that did not use any components of the NBCCEDP infrastructure, the Suffolk County, New York, site, created an effective collaboration between the existing infrastructures of an academic health center, a network of county health department–funded community health centers, and a program-funded patient navigator to bring patients into and through the screening process and back to their primary care setting.[5, 10]
Partnerships facilitated early through late CRCSDP implementation, aiding start-up through sustainability planning.[8, 13, 16, 19] Sites invested time and resources into relationship building that yielded critical clinical partnerships,[9, 16] a greater reach to priority groups, and tangible resources, including those for cancer treatment.[9, 19] Partners served as champions and change agents, promoting the programs, enhancing program legitimacy with the public, public health and medical communities, and patients, and communicating early program successes to key stakeholders.[8, 19] Community health centers and Federally Qualified Health Centers (FQHCs) were important clinical partners for several programs.[5, 9, 11, 12] The Suffolk County, New York, program established a system in which routine primary care before and after the procedure was delivered by the community health center with a seamless referral made into the medical center for the screening test.[8, 10] The St. Louis program worked to systematize patient referral from local FQHCs into the program-funded colonoscopy screening site, facing some challenges forging this collaboration at a grassroots level with limited FQHC infrastructure to support providing a new service. Other key partners included CDC-funded state-based National Comprehensive Cancer Control Programs and their coalitions, local chapters of the American Cancer Society, medical advisory boards, and universities.[8, 9, 11, 12, 16, 19]
Patient recruitment was identified as a challenge during early implementation. Primary dependence by the programs on in-reach to clients of the NBCCEDP, reluctance by programs to overrecruit early on, and limited resources dedicated to public education and outreach may have contributed to low screening numbers in the first full implementation year.[2, 12] Sites responded to slow early recruitment by developing more tailored outreach and education, some specifically geared toward men, partnering with existing referral services, expanding partnerships with primary care networks, improving and systematizing the patient enrollment process, and broadening and/or changing the screening test options.[2, 9, 11, 12]
The well-defined patient pathways, clinical protocols, quality assurance, and tracking systems to guide referral through diagnostic follow-up were critical facilitators in mid- and late implementation. Protocols that could be implemented with limited involvement of the provider, allowing the provider to focus on service delivery, were effective, for example, using medical assistants to flag charts of patients eligible for screening and using facsimiles to transmit colonoscopy referrals from participating clinics into screening centers.[9, 10] Of note, even with the ability to build these improvement systems and protocols into preexisting initiatives, up-front time-intensive technical assistance was required.[8, 12]
Clinical Service Delivery Challenges and Successes
Test performance characteristics and quality
Data analysis methods for the clinical outcomes and test performance are fully described in several articles in this Supplement.[2, 3] Overall, among the 5233 persons screened through this program (2294 using screening fecal occult blood tests [FOBTs] and 2937 using screening colonoscopies), 752 adenomatous polyps and 18 cancers were detected (24% and 0.6% of the total number of people screened, respectively). Treatment was initiated for all patients in whom cancers were diagnosed. The colonoscopy complication rate was low (2.38/1000), with no deaths reported.
Three programs—Nebraska, St. Louis, and Suffolk County, New York—chose FOBT as the primary screening test. An FOBT is an effective screening test only if patients return their test kits, receive diagnostic follow-up colonoscopy for a positive test, and undergo regular rescreening (annual or biennial). The CRCSDP programs that screened with FOBT faced challenges in all these areas, with low card return rates (53%), lower-than-desired rates of follow-up testing for positive FOBT tests (84%), and low rates of rescreening (13%-16%).[2, 3] Low rescreening rates may have been partially a result of the relatively short 4-year program period and the scheduled closeout of screening in the final program year.
As expected, given the inherent differences in the 2 tests, we observed wide variability in test performance between the FOBT and colonoscopy. Among average-risk persons, polyp, adenoma, and cancer detection rates were 5-10 times lower for those tested with a single guaiac-based FOBT (29.3, 18.5, and 1.8, respectively) compared with screening colonoscopy (370.6, 251.7, and 5.0, respectively). The lost-to-follow-up rate during the screening and follow-up cycle was higher with FOBT than colonoscopy (16% vs 1.5%).
Given the increasing attention to the quality of the colonoscopy performance,[20-23] many recently developed colonoscopy performance indicators were measured during the CRCSDP evaluation. The method for assessing these performance indicators as well as a comparison of CRCSDP quality assessment with assessments by other programs is more fully described in an accompanying article by Nadel et al. Among key quality indicators, CRCSDP cecal intubation rates (>90% at all 5 sites) and adenoma detection rates (21% for women and 32% for men across all 5 sites) reached recommended target levels suggested by expert groups, but rescreening and surveillance intervals following colonoscopy often did not concur with national guidelines.[3, 20, 22, 24] In addition, several key colonoscopy performance indicators, including bowel preparation quality and completeness of polyp removal, were poorly documented. It is likely that compliance with these reporting and practice guidelines is even lower outside organized program settings like that of the CRCSDP, with its established policies, oversight, and nonclinical support services.
Comparative clinical and nonclinical costs
Methods for the economic evaluation of the CRCSDP have been fully described in accompanying articles by Subramanian and Tangka[6, 7] and included cost analysis of both clinical (delivery of colorectal cancer screening and diagnostic services, with associated office visits and laboratory fees) and nonclinical (program management, patient support/patient navigation, public education and outreach, quality assurance and professional development, partnership development and maintenance, data collection and tracking, and program evaluation) activities. Of the nearly $10.5 million expended by the 5 sites on clinical and nonclinical services over 4 years, 71% of funding was provided by the CDC, with the remaining funds provided as in-kind contributions from individual sites. The proportion of funding spent on clinical service delivery and service delivery/patient support (eg, patient navigation) increased over the years from 52% in year 2 to 63% in year 4 and comprised the largest proportion of cost during the implementation phase (years 2-4). The average program cost per person varied by year, site, and test choice. In general, the larger was the number of persons screened by the site, the lower was the average program cost per person.
Based on a single screening cycle for FOBT, screening with FOBT, including colonoscopy follow-up, was less costly than colonoscopy, both per person and per program. The total average clinical cost per patient screened by FOBT (including costs for screening, diagnosis, initial surveillance, office visits, and associated clinical services averaged across all persons receiving screening FOBTs) ranged from $47.52 in Nebraska to $149.02 in greater Seattle. This compared with an average per-person clinical cost for all services related to colonoscopy screening ranging from $654.37 in St. Louis to $1599.79 in Baltimore city. The per-unit procedure cost charged by some providers varied from local Medicare rates. Data from the CRCSDP did not allow us to compare costs of an FOBT program with a colonoscopy program over multiple years, during which annually repeated FOBTs and diagnostic colonoscopies would accrue.
These variations in clinical test costs reflected differences in how each program was organized, which services were provided, and how a program paid for clinical tests and services. Programs attempted some measures of cost containment by negotiating prices, purchasing screening tests in bulk, or limiting reimbursement to certain components of the screening cycle. Variability in provider reimbursement reported for the FOBT tests was a result in part of differences in the bundling of services across programs. Providers were reimbursed for the FOBT kit alone, or it was bundled with the laboratory charge, the staff time devoted to patient coordination (including data collection, enrollment, tracking, and facilitating follow-up) or the cost of the office visit. As described by Subramanian et al, a substantial proportion of costs to the program was incurred by patient support and nonclinical activities.
Moving Forward—A Strengthened Population-Based Approach
Based on overall evaluation findings, we have demonstrated that it is feasible to establish a publically funded colorectal cancer screening program. Multiple models were viable, with clear benefits of building on and integrating into existing chronic disease programs. However, the overall programmatic model used in the CRCSDP, focused on funding screening services, limited the number of people reached, leading to the development of a subsequently more population-based program.
The CDC's second-generation program, the CRCCP, with its ambitious goal of increasing colorectal cancer screening prevalence to 80% in funded states and tribal areas by the end of the program period (2014), adopted a different program design to place greater emphasis on a population-based approach. The CRCCP, currently in 25 states and 4 tribal organizations, includes 2 program components: 1) screening promotion, involving activities to encourage increased population-level screening; and 2) screening provision, providing funding for clinical service delivery for low-income, underinsured persons. The lessons learned from the CRCSDP regarding implementation, clinical service delivery, and cost informed the design of the CRCCP.
Building on efforts begun in the demonstration program, the CRCCP places increased emphasis on evidence-based activities aimed at increasing population-level use of CRC screening. A socioecological model was adopted for the CRCCP including recommended systems and community- and policy-level strategies rather than individual-level strategies intended to maximize impact. Current grantees use evidence-based strategies recommended by the Guide to Community Preventive Services to inform their interventions, including client reminders, small media, reducing structural barriers, provider reminders, and provider assessment and feedback.
Strategic partnerships established in the CRCSDP and aimed at increasing the use of high-quality clinical preventive services across existing health systems have been strengthened. Like the CRCSDP sites, the CRCCP sites partner with health systems, private and public insurers, and FQHCs and encourage the increased use of client and provider reminders at the organizational level to maximize population reach. The CRCCP sites will continue to leverage traditional and innovative partnerships across other US Department of Health and Human Services (HHS) agencies and with existing health systems to maintain momentum on increasing CRC screening. Initiatives like Million Hearts, co-led by 2 HHS agencies, the CDC and the Centers for Medicare and Medicaid Services, which efficiently bridges clinical and community interventions to reduce heart attacks and strokes, may offer another important model for consideration that may be relevant to efforts to increase US colorectal cancer screening rates.
Clinical Service Delivery
As with the demonstration program, the CRCCP offers direct screening services to under- or uninsured men and women using any of the US Preventive Services Task Force recommended CRC screening tests. Measures were taken in this newer program to address the challenges faced in the CRCSDP. CRCCP grantees using stool testing have invested appropriate resources including reliable and systematized processes to assure that FOBT cards are returned, followed up with appropriate testing (usually colonoscopy), and repeated as recommended. Strong emphasis has been placed on the use of reminder systems, clinical decision support tools, and patient navigation to help guide test completion for each step of the screening cascade for any of the recommended screening tests. Guidance was also provided on recommended fecal occult blood tests.
Analysis of CRCSDP data identified the need for improved documentation of important elements in the colonoscopy report, the need to emphasize appropriate intervals for rescreening and surveillance intervals to minimize risk and better manage resources, and a continued need for a focus on systems and performance measurement to assure that the quality of the clinical service delivery is high.[2, 3] Based on analysis of CRCSDP clinical data, the CDC continues to encourage screening programs and clinical communities to conduct routine monitoring of colorectal cancer screening quality indicators as part of an ongoing quality improvement system. In the current CRCCP, CDC and site staff monitor quality indicators biannually.
CRCSDP sites that used the FOBT as their primary test were able to reach more clients at a lower cost after a single round of screening, but with screening test performance 5-10 times lower for the first FOBT than for colonoscopy.[6, 7] In comparison, based on current reimbursement rates for colonoscopy compared with FOBT, a much smaller cohort of individuals can be screened by colonoscopy in a given year than with the lower-cost FOBT with a similar program budget. It is important to note, however, that these comparisons are based on a short 4-year program without the benefit of repeated testing over time. Both test performance and costs are likely to be higher for the FOBT over the life of a program using repeated FOBTs annually or biennially. Differences in test performance and cost between FOBT and colonoscopy screening are also expected to narrow over time. Additional economic evaluation of the CRCCP may assist in examining these issues. A recent study that used a modeling approach to compare lifetime cost and benefits of FOBT versus colonoscopy concluded that although colonoscopies are more accurate, high-sensitivity FOBTs may result in more efficient allocation of available program funds. Large-scale colorectal cancer screening implemented in the future might consider potential efficiency gained from using high-sensitivity FOBTs or tests with similar effectiveness and cost. Regardless of which test is selected, the relatively high cost to screen per person and limited program reach contributed to the shift to a population-based approach in the CRCCP.
CRCSDP data pointed to the strengths and limitations of both the FOBT and colonoscopy, and data are continuing to accumulate about the potential benefits of other screening tests. Based on the evaluation of this program, we observed challenges associated with both FOBT (substantial loss to follow-up) and colonoscopy (higher program costs translating into fewer people screened). There may be compelling reasons to consider specific tests in certain settings, but CRC screening test choice continues to be important for clinicians, patients, health care systems, insurers, and a variety of other stakeholders in the United States. In addition, issues such as capacity, patient and provider preference, cost, and test performance need to be considered by programs and communities.[32, 33] Test-specific considerations observed from this initiative and relevant to future colorectal cancer screening efforts include return rates, diagnostic follow-up rates, and rescreening rates in a programmatic setting using the FOBT and quality, scalability, and cost in a wide-scale effort to use colonoscopy.
Recent increases in CRC screening have been driven by colonoscopy, whereas the use of FOBT has decreased over the last 4 years. Given current limits on colonoscopy capacity for some regions and populations and variations in patient acceptability, FOBT utilization must also be increased to maximize screening rates. CRCSDP programs encountered problems with patient and provider perceptions that FOBT is an inferior test, limiting its acceptance as a viable alternative to colonoscopy. Lessons can be learned from other countries and from some managed care systems in the United States that rely primarily on FOBT and that have developed strong systems to educate the public about available screening options, to identify eligible members to participate in recommended screenings, and to implement processes to assure appropriate follow-up of positive tests. Grantees in the CRCCP that are using the FOBT, particularly the immunochemical FOBT, will also contribute critical lessons learned to further inform effective ways to reach the population and reduce the burden of this largely preventable disease.
Future evaluations of new, innovative programs like the CRCSDP should consider developmental evaluation, an approach that facilitates ongoing program modification and adaptation in complex environments or systems. The CRCSDP evaluation produced some important products available to others. These include 2 data sets—one to monitor clinical service delivery and another to assess costs by program component—both being used in the new CRCCP. As others use these tools in evaluating their own programs, opportunities for comparison or meta-analysis with CRCSDP data may be possible. Overall, the small sample size of the CRCSDP limited our ability to evaluate population-level impact and cost effectiveness.
Broader population-based approaches such as those implemented in the CRCCP that link elements of an organized screening program into health systems are still needed. An organized screening program is defined by the International Agency for Research on Cancer as having the following characteristics: 1) an explicit policy with specified age categories, method, and interval for screening; 2) a defined target population; 3) a management team responsible for implementation; 4) a health care team responsible for decisions and care; 5) a quality assurance structure; and 6) a method for identifying cancer occurrence in the population. The system-level changes being accomplished through the CRCCP will complement other ongoing initiatives such as Medicaid expansion, subsidized state insurance exchanges, and elimination of cost sharing. However, as with the CRCSDP, it is likely that removing financial barriers will be necessary but not sufficient to recruit patients for screening. To realize the full potential of anticipated improvements in access to care, tailored outreach and education are necessary to reach underserved and all eligible populations. Further, systems-level interventions are needed to implement evidence-based approaches in community and clinical venues to routinize the use of colorectal cancer screening and other preventive services as a part of overall care.
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.
CONFLICT OF INTEREST DISCLOSURES
The authors made no disclosures.
- 14Framework for program evaluation in public health. MMWR Recomm Rep. 1999;48:1–40.
- 16Start-up of the Colorectal Cancer Screening Demonstration Project. Prev Chronic Dis. 2008;5:A38., , , , , .
- 17National Breast and Cervical Cancer Early Detection Program. Atlanta, GA: Centers for Disease Control and Prevention. http:www.cdc.gov/cancer/nbccedp. Accessed March 31, 2012.
- 18WISEWOMAN—Well-Integrated Screening and Evaluation for Women across the Nation. Atlanta, GA: Centers for Disease Control and Prevention. http:www.cdc.gov/wisewoman. Accessed March 31, 2012.
- 19Facilitators and challenges to start-up of the Colorectal Cancer Screening Demonstration Program. Prev Chronic Dis. 2008;5:A39., , , , .
- 20Quality in the technical performance of colonoscopy in the continuous quality improvement process for colonoscopy: Recommendations of the Multi-Society Task Force on Colorectal Cancer. Am J Gastroenterol. 2002;97:1294–1308., , , et al.Direct Link:
- 26Guide to Community Preventive Services. Atlanta, GA: Centers for Disease Control and Prevention. http://www.thecommunityguide.org/about/task-force-members.html. Accessed March 31, 2012.
- 27Colorectal Cancer Control Program. Atlanta, GA: Centers for Disease Control and Prevention. http://www.cdc.gov/cancer/crccp/. Accessed April 10, 2013.
- 28Million Hearts. U.S. Department of Health and Human Services. http://millionhearts.hhs.gov/index.html. Accessed March 29, 2011.
- 29U.S. Preventive Services Task Force. Screening for Colorectal Cancer: Clinical Summary of U.S. Preventive Services Task Force Recommendation 2008. http://www.uspreventiveservicestaskforce.org/uspstf08/colocancer/colosum.htm. Accessed February 24, 2011.
- 37Developmental Evaluation: Applying Complexity Concepts to Enhance Innovation and Use. New York, NY: Guilford Press; 2011..
- 39International Agency for Research on Cancer. Cervix cancer screening. IARC Handbook of Cancer Prevention. Lyon, France: IARC; 2005:10:117–162.