Adenoma miss rate determined by very shortly repeated colonoscopy: Retrospective analysis of data from a single tertiary medical center in China

Adenoma miss rate (AMR) has been calculated in several tandem colonoscopy studies, but it costs overmuch to carry out a clinical trial. We aimed to put forward AMR by taking advantage of retrospective data, and to judge the comparability between AMRs from prospective and retrospective data. Data of the patients accepting repeated colonoscopies during January to September 2016 was retrospectively collected and analyzed. Information was recorded, including bowel preparation quality of the first colonoscopy, size, location, histology and whether missed within the first colonoscopy of each single adenoma. AMR was compared by different risk factors through χ(2) test and multivariable logistic regression. Around 267 adenomas were detected during 309 pairs of repeated colonoscopies, of which 66 were missed during the first colonoscopies. AMRs of the lesions small in size, nonadvanced in histology, in poor bowel preparation context and located in the proximal colon, were significantly higher than the opposite ones, and old age and male were related to adenoma missing (P < .05). In multivariable logistic regression analysis, adenoma-related factors (diminutive in size, poor bowel preparation and located in ascending colon, transverse colon or sigmoid colon), and patient-related factors (older than 60 years, male and poor bowel preparation) were found to be independently associated with missing adenomas (P < .05). AMR of retrospective data is comparable to that of tandem studies. Several risk factors influence AMR dramatically, which should be paid attention to.