Air Insufflation vs Water Exchange Colonoscopy: A Multiple Cross-over, N-of-1 Study in a Trainee

Abstract

Listen

Translate article

Introduction: Water exchange (WE) colonoscopy is associated with less discomfort when compared to air insufflation (AI); other likely advantages include a lesser occurrence of loop formation and a higher adenoma detection rate. Yet despite these advantages, colonoscopy training in most endoscopy programs starts with AI. Aim: To study outcomes in a trainee who was randomly assigned to AI vs. WE during the first two weeks of colonoscopy training. Methods: For AI the trainee followed the advice of various staff members; there was no protocol. For WE the trainee worked with a single endoscopist who instructed the following method: (1) all air was removed during insertion and water was used to identify proximal colon; (2) stool containing water was exchanged for clean water; (3) water volume was maintained at the lowest amount allowing endoscope advancement; (4) mandatory rotation of the patient from left to back at the splenic flexure; (5) manual pressure with splenic and hepatic flexure passage. Data collected included gender, age, sedation, insertion extent and time, total time, loop formation, number and type of polyps; T-test was used for analysis. Results: The trainee performed 35 procedures in the first 2 weeks of training with multiple cross-overs: 3AI - 3WE - 4AI - 4WE - 3AI - 4WE - 10AI - 4WE. Gender distribution was 4/20 for AI and 4/15 for WE; ages were 65±9 year (M±SD ) for AI and 60±7 for WE (p=0.15). For AI cecal intubation was successful in 12/20 (60%) patients during 21±9min; for WE this was 14/15 (93%) during 9±3min (p < 0.001). The terminal ileum was intubated in 1/20 during AI and 10/15 during WE. Significant loop formation occurred in 7/20 during AI and 1/15 during WE. For AI sedation consisted of fentanyl 96±32mcg and midazolam 4.5±1.8mg; for WE this was 57±20mcg and 2.5±0.6mg (p < 0.001 both). The ADR was 0.5 during AI and 0.66 during WE (p=.33). Total endoscopy time was 47±18min for AI vs. 42±6min for WE (p=.44). Limitations: This is the experience of a single trainee. The WE technique was more structured than the AI technique. During later AI colonoscopies the trainee started using WE techniques given the experience with WE. Conclusion: WE is statistically superior to AI for most features that were measured for this starting trainee: insertion time, cecal and terminal ileum intubation rates, prevention of loop formation and need for sedation. WE had a trend towards a higher ADR and shorter total endoscopy time. WE should be considered as starting technique for colonoscopy training.