Nine randomised trials were included (623 participants). Sample sizes ranged from 27 to 131. Three trials assessed flexible sigmoidoscopy. Six trials assessed colonoscopy. The authors stated insufficient reporting of methodological details in most trials, especially for generation of allocation sequence, allocation concealment and power calculations. In most trials patients and/or outcome assessors were blinded to the intervention. Six trials used intention-to-treat analysis.
Patient experience: Three colonoscopy trials reported no differences in pain scores between the N2O group and control for pre-procedural anxiety. Six studies reported on procedural pain. One colonoscopy study reported significantly lower pain scores for the N2O group compared to control (16.7 points and 40.1 points) and one colonoscopy study reported higher pain scores in the N2O group compared to control (33 points and 3 points).
The other trials (three flexible sigmoidoscopy and one colonoscopy) reported no significant differences between groups. One flexible sigmoidoscopy trial reported lower levels of procedural discomfort in the N2O group compared to control (30 points and 71 points). Two trials (one flexible sigmoidoscopy, one colonoscopy) reported no significant differences between groups for procedural discomfort. No significant differences were reported for the for extra intravenous sedation between groups in five colonoscopy trials.
Adverse effects (five trials): One colonoscopy trial reported more adverse events in the N2O group (15 events) compared to control (five events). Four trials (one sigmoidoscopy and three colonoscopy) reported no significant differences between groups.
Psychomotor recovery (four trials): Two colonoscopy trials reported better psychomotor recovery in the N2O group compared to control immediately after procedures. One colonoscopy trial reported mixed results for psychomotor recovery. Results were not reported for one trial.
Length of post-procedural stay (seven trials): Five colonoscopy trials and one sigmoidoscopy trial reported significantly shorter time-periods for post-procedural stay for N2O. One colonoscopy trial reported no differences between groups.