* Setting: Screening colonoscopy or secondary colonoscopy after positive FOBT performed in endoscopic unit of a single university hospital
* Study design: Single blinded randomized trial
* Randomization process: sealed envelope
* Study subjects:
Inclusion criteria: Subjects who receive Screening colonoscopy or secondary colonoscopy Exclusion criteria: subjects who have COPD or cardiovascular diseases
* Study duration: June 2008 to November 2012
* Post-colonoscopy discomfort will be measured and obtained from the screenee at several time points using visual analogue pain scoring system:
* Immediately after colonoscopy for non-sedated patients, immediately after awaken from sedation for sedated patients
* 2 hrs after colonoscopy
* 8-12 hours after colonoscopy
* Statistical analysis:
* Sample size estimation: according to our previous unpublished data using air insufflations at colonoscopy, mean pain score was 3 and standard deviation was 2.38. Assuming 15% difference of discomfort, sample size up to 171 are required for each group (CO2 vs. air, sedated and non-sedated) in this study. (2 sided test, α=0.05, power=80%)
* Student t-test for pain score comparison
* Ancillary analysis: pCO2, sedation drug dosage, and procedural time in both groups. For TCCO2 comparison, ANOVA for repeated measurement will be used. For variables not normally distributed, Wilcoxon rank sum test wall be used as a supplementary analysis.
* SAS statistical program, version 9.0 (SAS institute Inc., Cary, NC) will be used for all statistical analysis
