A proof-of-principle, prospective, randomized, controlled trial demonstrating improved outcomes in scheduled unsedated colonoscopy by the water method

BackgroundAn observational study in veterans showed that a novel water method (water infusion in lieu of air insufflation) enhanced cecal intubation and willingness to undergo a repeat scheduled unsedated colonoscopy.ObjectiveTo confirm these beneficial effects and significant attenuation of discomfort in a randomized, controlled trial (RCT).DesignProspective RCT, intent-to-treat analysis.SettingVeterans Affairs ambulatory care facility.PatientsVeterans undergoing scheduled unsedated colonoscopy.InterventionsDuring insertion, the water and traditional air methods were compared.Main Outcome MeasurementsDiscomfort and procedure-related outcomes.ResultsEighty-two veterans were randomized to the air (n = 40) or water (n = 42) method. Cecal intubation (78% vs 98%) and willingness to repeat (78% vs 93%) were significantly better with the water method (P < .05; Fisher exact test). The mean (standard deviation) of maximum discomfort (0 = none, 10 = most severe) during colonoscopy was 5.5 (3.0) versus 3.6 (2.1) P = .002 (Student t test), and the median overall discomfort after colonoscopy was 3 versus 2, P = .052 (Mann-Whitney U test), respectively. The method, but not patient characteristics, was a predictor of discomfort (t = −1.998, P = .049, R2 = 0.074). The odds ratio for failed cecal intubation was 2.09 (95% CI, 1.49-2.93) for the air group. Fair/poor previous experience increased the risk of failed cecal intubation in the air group only. The water method numerically increased adenoma yield.LimitationsSingle site, small number of elderly men, unblinded examiner, possibility of unblinded subjects, restricted generalizability.ConclusionsThe RCT data confirmed that the water method significantly enhanced cecal intubation and willingness to undergo a repeat colonoscopy. The decrease in maximum discomfort was significant; the decrease in overall discomfort approached significance. The method, but not patient characteristics, was a predictor of discomfort. (Clinical trial registration number NCT00747084).