Evaluating Test Strategies for Colorectal Cancer Screening—Age to Begin, Age to Stop, and Timing of Screening Intervals

Background: The U.S. Preventive Services Task Force requested a decision analysis to inform their update of the recommendations for colorectal cancer (CRC) screening. Objective: To assess life-years gained and colonoscopy requirements for CRC screening strategies and identify a set of recommendable screening strategies. Design: Decision analysis using two CRC microsimulation models from the Cancer Intervention and Surveillance Modeling Network. Data Sources: Derived from recent published literature on test characteristics of single use applications of various screening strategies. Target Population: U.S. average-risk 40-year-old population. Perspective: Societal. Time Horizon: Lifetime. Interventions: Fecal occult blood tests (FOBTs), flexible sigmoidoscopy, or colonoscopy screening beginning at age 40, 50, or 60 and stopping at age 75 or 85 with screening intervals of 1, 2, or 3 years for FOBT and 5, 10, or 20 years for sigmoidoscopy and colonoscopy. Outcome Measures: Number of life-years gained compared with no screening and number of colonoscopies and non-colonoscopy tests required. Results of Base-Case Analysis: Beginning screening at age 50 was consistently better than age 60. Lowering the stop age from 85 to 75 decreased life-years gained by 1% to 4%, while colonoscopy use fell by 4% to 15%. Assuming equally high adherence, four strategies provided comparable life-years gained, namely 10-yearly colonoscopy, annual Hemoccult SENSA or fecal immunochemical test, and 5-yearly flexible sigmoidoscopy in conjunction with Hemoccult SENSA every 2 to 3 years. Annual Hemoccult II alone and 5-yearly flexible sigmoidoscopy alone were less effective. Results of Sensitivity Analysis: The results were most sensitive to beginning screening at age 40. Limitations: Stopping age for screening was based only on chronological age. Conclusions: Our findings support CRC screening from ages 50 to 75 with annual screening with a high sensitivity FOBT, 10-yearly colonoscopy, or high sensitivity FOBT every 2 to 3 years with a 5-yearly flexible sigmoidoscopy. INTRODUCTION Despite recent declines in both incidence and mortality (1), colorectal cancer (CRC) remains the second most common cause of cancer death in the United States (2). Screening for CRC reduces mortality through the detection of malignancies at earlier, more treatable stages, as well as through the identification and removal of adenomatous polyps (asymptomatic benign precursor lesions that may lead to CRC). There are a number of tests currently available for screening, such as fecal occult blood testing (FOBT), flexible sigmoidoscopy, and colonoscopy. Screening with FOBT (Hemoccult II) has been shown to reduce CRC mortality by 15% to 33% in randomized controlled trials (3-5) and screening with more sensitive FOBTs, flexible sigmoidoscopy, colonoscopy or combinations of these tests may reduce the burden of CRC even more (6, 7). In the absence of adequate clinical trial data on several recommended screening strategies, microsimulation modeling can provide guidance on the risks, benefits, and testing resources required for different screening strategies to reduce the burden of CRC. In July 2002, the US Preventive Services Task Force (USPSTF) concluded that there was sufficient evidence to recommend strongly that all average-risk adults 50 years of age and older should be offered CRC screening (8). However, the logistics of screening such as the type of screening test, screening interval, and age to stop screening were not evaluated in terms of the balance of benefits and potential harms. The USPSTF has again addressed CRC screening recommendations with a systematic review of the evidence (9, 10) on screening tests. For this assessment, the Task Force requested a decision analysis to project expected outcomes of various CRC screening strategies. Two independent microsimulation modeling groups from the Cancer Intervention and Surveillance Modeling Network (CISNET), funded by the National Cancer Institute, used a comparative modeling approach to compare life-years gained relative to resource use of different CRC screening strategies. The report to the United States Preventive Services Task Force was published in November 2008 (11). This is a fuller version of that report and includes additional tables and methodological descriptions not included in the publication in Annals of Internal Medicine. METHODS We used two microsimulation models, MISCAN (MI-crosimulation Screening ANalysis) (12-14) and SimCRC (Simulation Model of Colorectal Cancer) (15), to estimate the life-years gained relative to no screening and the colonoscopies required (i.e., an indicator for resource use and risk of complications) for different CRC screening strategies defined by test, age to begin screening, age to stop screening, and screening interval. We aimed to identify a set of recommendable strategies with comparable clinical benefit and an efficient use of colonoscopy resources. Using two models (i.e., a comparative modeling approach) adds credibility to the results and serves as a sensitivity analysis on the underlying structural assumptions of the models, particularly pertaining to the unobservable natural history of CRC. Microsimulation Models A detailed description of the MISCAN and SimCRC models can be found in Appendix 1. Standardized model profiles are available online (http://cisnet.cancer.gov/profiles/), which