Objective: Studies of tandem colonoscopies have reported that 15-25% of neoplastic polyps <1 cm in size and 0-6% of larger polyps are overlooked at the time of colonoscopy. We determined the miss rate and "true" 1-yr recurrence of neoplastic polyps in a population of patients reflecting a broad spectrum of different gastroenterology practice settings.
Methods: Patient data from several sources were examined for repeat colonoscopies performed on the same patient within 120 days of each other. Examination pairs were included for analysis if both colonoscopies had good preps and reached the cecum. The miss rate was calculated by two methods: 1) a pooled rate, the total number of polyps on all second examinations divided by the total number on both examinations, and 2) a within-person rate, the average of the individual miss-rates. We estimated a "true" 1-yr recurrence rate by subtracting the proportion of patients with a missed neoplastic polyp from the proportion of patients with a neoplastic polyp found at 1 yr.
Results: A total of 76 colonoscopy pairs a mean 47 days apart (range, 1-119 days) were identified from a total of approximately 15,000 examinations and used to calculate the overall miss rates. For the category "all polyps" (neoplastic and nonneoplastic polyps), 17% by the pooled method and 11% by the within-person method were missed. The corresponding rates for neoplastic polyps were 12% by the pooled method and 8% by the within-person method. A total of 17% of patients had one or more neoplastic polyps missed on the initial examination. The observed 1-yr recurrence rate was determined from 1,314 colonoscopy pairs performed a mean of 379 days apart and found to be 28% for neoplastic polyps. By comparing this to the proportion of patients with one or more missed neoplastic polyps, we found the true 1-yr recurrence of neoplastic polyps to be 11%.
Conclusion: There is a significant colonoscopic miss rate for neoplastic polyps and "all polyps" in clinical practice that is comparable to that previously stated in special research settings. The within-person rate more accurately reflects the true colorectal polyp miss rate for any given colonoscopic exam than does the pooled rate. The difference between the observed 1-yr recurrence rate and the proportion of patients with a missed polyp represents the true 1-yr recurrence of neoplastic polyps.