Adherence to lower endoscopic surveillance in high-risk colorectal cancer populations: A meta-analysis of U.S. studies.

Abstract

e22536 Background: High-risk individuals, including those with a history of colorectal cancer (CRC), advanced adenomas, or hereditary syndromes, require routine lower endoscopic surveillance to reduce recurrence and mortality. However, real-world adherence in U.S. healthcare systems remains poorly characterized. This meta-analysis quantified adherence and examined variability by surveillance interval ( < 2 vs. ≥2 years). Methods: We conducted a systematic review and random-effects meta-analysis of studies reporting surveillance adherence among high-risk CRC populations in U.S. healthcare settings. Eligibility was restricted to peer-reviewed studies published in English that reported quantifiable adherence rates to lower endoscopic surveillance (e.g., colonoscopy, CT colonography). Subgroup analyses stratified studies by surveillance interval ( < 2 years vs. ≥2 years post-index). Pooled estimates were calculated using a logit transformation. Between-study heterogeneity was assessed using I² and τ², and prediction intervals were reported to estimate expected adherence in future settings. Results: The primary meta-analysis (4 studies; N = 11,071) demonstrated a pooled adherence rate of 44% (95% CI: 19–72%; prediction interval: 5–92%; I² = 99.7%). Subgroup analyses, which included additional eligible studies contributing interval-specific data, showed higher adherence in the < 2 years surveillance interval (3 studies; N = 16,884; 56%; 95% CI: 3–98%; I² = 99.9%). In the ≥2 years subgroup (5 studies; N = 12,305), pooled adherence was 45% (95% CI: 19–74%; I² = 98.8%). Considerable between-study heterogeneity was observed across analyses. Conclusions: Adherence to lower endoscopic surveillance among high-risk CRC populations in the U.S. remains suboptimal and variable. Shorter surveillance intervals were associated with modestly higher adherence, though estimates varied substantially across studies. These findings highlight the need for improved surveillance tracking systems and targeted interventions. An ongoing thematic synthesis will further characterize multilevel barriers to inform future quality improvement efforts.