Projected colorectal cancer incidence and mortality based on observed adherence to colonoscopy and sequential stool-based screening

Reinier G.S. Meester,Iris Lansdorp-Vogelaar,Sidney J. Winawer,Timothy R. Church,John I. Allen,Andrew D. Feld,Glenn Mills,Paul A. Jordan,Douglas A. Corley,Chyke A. Doubeni,Anne I. Hahn,Stephanie M. Lobaugh,Martin Fleisher,Michael J. O'Brien,Ann G. Zauber
DOI: https://doi.org/10.14309/ajg.0000000000002693
2024-02-07
The American Journal of Gastroenterology
Abstract:INTRODUCTION: Modeling supporting recommendations for colonoscopy and stool-based colorectal cancer (CRC) screening tests assumes 100% sequential participant adherence. The impact of observed adherence on the long-term effectiveness of screening is unknown. We evaluated the effectiveness of a program of screening-colonoscopy every ten years versus annual high-sensitivity guaiac-based fecal occult blood testing (HSgFOBT) using observed sequential adherence data. METHODS: MIcrosimulation SCreening ANalysis (MISCAN) model using observed sequential screening adherence, HSgFOBT positivity, and diagnostic-colonoscopy adherence in HSgFOBT-positive individuals from the National Colonoscopy Study (NCS; single screening-colonoscopy versus ≥4 HSgFOBT sequential rounds). We compared CRC incidence and mortality over 15 years with no screening, or ten-yearly screening-colonoscopy versus annual HSgFOBT with 100% and differential observed adherence from the trial. RESULTS: Without screening, simulated incidence and mortality over 15 years were 20.9 (95% probability interval, 15.8-26.9) and 6.9 (5.0-9.2) per 1000 participants, respectively. In the case of 100% adherence, only screening-colonoscopy was predicted to result in lower incidence; however, both tests lowered simulated mortality to a similar level (2.1 [1.6-2.9] for screening-colonoscopy; 2.5 [1.8-3.4] for HSgFOBT). Observed adherence for screening-colonoscopy (83.6%) was higher than observed sequential HSgFOBT adherence (73.1% first round; 49.1% by round 4), resulting in lower simulated incidence and mortality for screening-colonoscopy (14.4 [10.8-18.5] and 2.9 [2.1-3.9], respectively) than HSgFOBT (20.8 [15.8-28.1] and 3.9 [2.9-5.4], respectively), despite a 91% adherence to diagnostic-colonoscopy with FOBT positivity. The relative risk of CRC mortality for screening-colonoscopy versus HSgFOBT was 0.75 (95%PI, 0.68-0.80). Findings were similar in sensitivity analyses with alternative assumptions for repeat colonoscopy, test performance, risk, age, and projection horizon. DISCUSSION: Where sequential adherence to stool-based screening is suboptimal and colonoscopy is accessible and acceptable – as observed in NCS – offering screening-colonoscopy can increase screening effectiveness.
gastroenterology & hepatology
What problem does this paper attempt to address?