TOPLINE

In a large integrated health system, endoscopists adhered to the 2020 updated postpolypectomy surveillance recommendations in more than 90% of patients with three to four nonadvanced adenomas but in 68%-72% of those with one to two such adenomas. About 3 in 10 patients with one to two nonadvanced adenomas were assigned a surveillance interval shorter than recommended, suggesting the possibility of over-surveillance in this group.

METHODOLOGY

• In 2020, the US Multi-Society Task Force revised its postpolypectomy surveillance recommendations to 7-10 years for one to two nonadvanced adenomas and 3-5 years for three to four, but how closely endoscopists followed the update was unknown.
• Researchers conducted a cross-sectional study of 41,606 patients (mean age, 59.8 years; 57.6% men) at an average risk for colorectal cancer who underwent removal of one to four nonadvanced adenomas during baseline colonoscopy at an integrated health system between 2019 and 2024.
• An adenoma was considered nonadvanced if it was tubular, measured < 10 mm, and showed no high-grade dysplasia or villous histology; 33,787 patients had one to two nonadvanced adenomas and 7819 had three to four of these adenomas.
• A total of 281 endoscopists performed the baseline colonoscopies, with a median of 19 years in practice and an overall median adenoma detection rate of 52.2% in the prior year.
• The primary outcome was how often patients were assigned a surveillance interval within the range set by the 2020 update.

TAKEAWAY

• Endoscopists followed the recommended 7-10-year surveillance interval for 68.1% of patients with one to two nonadvanced adenomas in 2021. This rose to 71.9% in 2024. For patients with three to four nonadvanced adenomas, the recommended 3-5-year interval was followed in more than 90% of patients in every year from 2021 to 2024.
• Between 2021 and 2024, 28.1%-31.9% of patients with one to two nonadvanced adenomas were told to return for colonoscopy in under 7 years, and none were told to wait longer than 10 years. Among those with three to four nonadvanced adenomas, 2.3%-2.5% were told to return in under 3 years, and 6.6%-7.3% were told to wait longer than 5 years.
• For patients with one to two nonadvanced adenomas, nonadherent intervals were more likely with older age, male sex, BMI ≥ 30, smoking history, multiple comorbidities, and endoscopists with 15-24 years in practice, and less likely with diagnostic indications, Hispanic ethnicity, Asian race, and each additional year since the guideline update. No factors were significantly linked to nonadherence among patients with three to four nonadvanced adenomas.
• Endoscopists rarely chose the middle years of the recommended ranges, instead favoring the upper and lower bounds. For patients with one to two nonadvanced adenomas, they recommended 10 years in 45.9%-47.2% and 7 years in 19.8%-25.1%; for those with three to four nonadvanced adenomas, they recommended 3 years in 50.9%-51.5% and 5 years in 37.6%-39.5%.

IN PRACTICE

"The findings suggest that the potential for over-surveillance is substantially more likely for patients with 1-2 versus 3-4 non-advanced adenomas," the authors wrote.

"Guideline developers could consider recommending an interval of, for example, 7 or 10 years in place of the 7-10-year range and 3 or 5 years in place of the 3-5-year range," they added.

SOURCE

The study was led by Jeffrey M. Hendel, MD, Kaiser Permanente Medical Center, San Francisco. It was published online in Clinical Gastroenterology and Hepatology.

LIMITATIONS

Researchers drew the study population from a single integrated health system, which may have limited the generalizability of the findings to other practice settings. The observational study design may have left residual confounding and misclassification unaddressed. Assessment of adherence relied on adenoma findings at baseline colonoscopy and may not have accounted for all the clinical factors that shaped endoscopists' surveillance interval decisions.

DISCLOSURES

Study funding was not explicitly mentioned. The authors declared having no conflicts of interest.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.