Polyp detection rate in transverse and sigmoid colon significantly increases with longer withdrawal time during screening colonoscopy.

Background: The guidelines for colonoscopy present withdrawal time (WT) and adenoma detection rate (ADR) as the quality indicator. The purpose of this retrospective study is to analyze the predicting factors with polyp detection rate (PDR) as a surrogate for ADR by using comprehensive health checkup data, and assess the correlation between PDR per each colonic segment and WT, and factors influencing WT.
Methods: One thousand and thirty six consecutive health checkup cases from April 2015 to March 2016 were enrolled in this study, and 880 subjects who undertook colonoscopy without polyp removal or biopsy were divided into the two groups (polyp not detected group vs polyp detected group). The two groups were compared by subjects and clinical characteristics with univariate analysis followed by multivariate analysis. Colonoscopies with longer WT (≥ 6 min) and those with shorter WT (< 6 min) were compared by PDR per each colonic segment, and also by subjects and clinical characteristics.
Results: A total of 1009 subjects included two incomplete colonoscopies (CIR, 99.9%) and overall PDR was 35.8%. A multiple logistic regression model demonstrated that age, gender, and WT were significantly related factors for polyp detection (odds ratio, 1.036; 1.771; 1.217). PDR showed a linear increase as WT increased from 3 min to 9 min (r = 0.989, p = 0.000) and PDR with long WT group was higher than that with short WT group per each colonic segment, significantly in transverse (2.3 times, p = 0.004) and sigmoid colon (2.1 times, p = 0.001). Not only bowel preparation quality but also insertion difficulty evaluated by endoscopist were significant factors relating with WT (odds ratio, 3.811; 1.679).
Conclusion: This study suggests that endoscopists should be recommended to take more time up to 9 min of WT to observe transverse and sigmoid colon, especially when they feel no difficulty during scope insertion.