Reply

We thank Drs Hassan and Zullo for their interest in our article.¹ To our knowledge, there are no large cohort studies of long-term follow-up after screening colonoscopy to which our results can be directly compared. The value of colonoscopy as a screening tool rests in its ability to decrease colorectal cancer incidence and mortality at the population level, specifically, in persons with and without adenomatous polyps, curable stage adenocarcinoma, or both. Colonoscopy presumably benefits patients with adenomas by removal of a precancerous lesion; it identifies patients without adenomas for longer follow-up intervals; and it detects patients with curable cancer. Our aim was to present the effect of screening colonoscopy on colorectal cancer incidence and mortality in such a mixed cohort. In our report, we discuss the reasons for not presenting the results of the subgroup analysis conducted by Drs Hassan and Zullo. Our original intent was to report standardized incidence ratios (SIR) for the study cohort stratified by whether colon neoplasia was present at the baseline colonoscopy. This would have allowed a more direct comparison to other studies, particularly the National Polyp Study (NPS)² for the adenoma-bearing subgroup. However, the small sample size of the adenoma subgroup (n = 190, compared with 1418 in the NPS), and the rarity of incident cancers in the subgroup with no baseline adenomas (1 cancer) obviously precluded a robust analysis. The NPS shows how efficacious polypectomy can be, as opposed to how effective it is in the real world in which clearing exams may not be as fastidiously performed, or follow-up not as intense or as complete. Additionally, 13% of patients in the NPS underwent a second clearing colonoscopy, which is not done in usual practice, and patients with adenomas >3 cm were excluded. These considerations highlight major methodological differences between the NPS and subsequent studies, including our own.