Effectiveness of screening for colorectal cancer in average risk adults: Colonoscopy vs FIT

点击翻译按钮获取中文摘要

PROJECT SUMMARY: Cancers of the colon and rectum are the second leading cause of cancer deaths in the United States, but are potentially preventable through early detection of invasive or precancerous lesions. Of the currently recommended tests for screening average risk people, colonoscopy is the most commonly used in the United States, but stool-based tests such as fecal immunochemical tests (FIT), are the most commonly used globally. However, there is currently no strong evidence to inform highly consequential decisions on these competing screening strategies, which contributes to conflicting recommendations from policy groups including a Canadian task force recommendation against using colonoscopy for routine screening. Microsimulation studies report that FIT and colonoscopy can achieve similar overall lifetime effectiveness, but are based on assumptions of efficacy and adherence to regular screening that are not attainable in real practice. An added challenge is the paucity of well-developed statistical approaches that can account for varying levels of adherence, patterns of use such as cross-overs, and time-dependent confounding when estimating the effectiveness of colorectal cancer (CRC) screening tests. Our proposed series of studies aim to: 1) estimate the effectiveness of FIT in reducing the risk of dying from CRC when compared with no screening, and 2) compare the effectiveness of FIT in reducing the risk of dying from CRC relative to screening colonoscopy. We will also assess for potential effect modification by race/ethnicity, sex, and socioeconomic status. We will use a nested case-control design with about 2000 cases and 8000 controls between 2014 and 2018 in a historical cohort of about 2 million screening-eligible people receiving care at Kaiser Permanente Northern and Southern California. Cases for the study will be men and women who were 52-89 years old at the date of death from CRC, each matched within health plan to four randomly selected controls based on age, sex, enrollment history, and geographic region. As secondary methodological aims, we will: 1) apply novel statistical methods to estimate FIT-colonoscopy comparative effectiveness in varying real-life screening patterns including imperfect adherence and crossover from FIT to colonoscopy; and 2) assess the use of a case-cohort design to inform future studies. As an exploratory aim, we will characterize the resources required to screen with FIT or colonoscopy as inputs for future microsimulation studies. Our study sites provide the full continuum of cancer care ranging from screening through end of life, and provide access to information on CRC screening that can be linked to diagnosis and cause-specific mortality data on large numbers of people. We will identify receipt of screening colonoscopy and screening FIT in the 10-year window prior to the reference date, differentiating screening from diagnostic procedures with selected chart audits to confirm colonoscopy screening indications. This will be the first observational study to address these critical scientific, clinical, and policy questions. The project will be implemented by an experienced team of high-impact scientists.