Key points are not available for this paper at this time.
OBJECTIVES: Colonoscopy can be uncomfortable. To increase safety, there is a trend, in the UK, towards reduced sedative use. We aimed to determine factors predictive of discomfort during colonoscopy. METHODS: Prospectively recruited patients were asked to grade anticipated discomfort on a Numeric Rating Scale ranging from 0 to 10. Discomfort scores were recorded every 2 min during the procedure and during peaks of discomfort. An overall discomfort score was recorded. RESULTS: One hundred and nine patients 44 male, 65 female; median 61.5 (21-80) years were recruited. One hundred and three procedures were completed. Forty-five patients received midazolam median 2 (1.5-5) mg. Mean overall Numeric Rating Scale score was 4.7 (men 4.0; women 5.2; P<0.01) and median peak score 7. Discomfort was usually greatest at the beginning of the procedure, while in the sigmoid colon. Discomfort scores were higher in patients with irritable bowel syndrome (P = 0.03); diverticular disease (P<0.01); midazolam (P = 0.02), buscopan (P<0.001) or nitrous oxide (P<0.001) use; endoscope tracker use (P = 0.01); incomplete procedures (P<0.001) or a preceding gastroscopy (P = 0.02), but were not correlated with discomfort during venous cannulation or digital rectal examination. Multivariate analysis showed that female sex, high anxiety, anticipation of high discomfort, longer intubation time and higher endoscopist's grade of procedural difficulty were independent factors significantly related to overall discomfort scores. Recollected discomfort scores 2-3 months later were lower (P<0.01). Low-dose midazolam had no appreciable amnesic effect. CONCLUSION: Factors indicative of difficult colonoscopy, and preceding gastroscopy, are associated with greater discomfort, as are the presence of female sex, irritable bowel, anxiety and anticipated discomfort. Low-dose midazolam neither relieves discomfort nor makes patients forget it. Selected patients may benefit from increased analgesia.
Elphick et al. (Fri,) studied this question.