Background: Since its introduction in the late 1960s, the technology of colonoscopy has developed rapidly, and the competence of the endoscopists has increased. Still, a complete colonoscopy is not always possible to perform.
Aim: To assess, in a population-based setting, which patient factors influence the success rate.
Patients and methods: All colonoscopy records during 1979-1995 in one Swedish county (population 258,000) were retrieved. Information was obtained about each patient's sex and age, date of examination, the endoscopist and indications for colonoscopy, findings, type of colonoscopy and reasons for an incomplete colonoscopy. Results were assessed by univariate and multivariate analyses.
Results: Of 5145 colonoscopies, 4153 (81%) were complete. Completion rates were influenced by (odds ratio for completion [95% confidence intervals]): sex, male versus female (2.00 [1.39-2.86]); age, youngest quartile versus oldest quartile (1.49 [1.04-2.13]); indication for colonoscopy, previous colonic surgery or long-standing colitis versus other indications (2.08 [1.23-3.45]); and diverticulosis, presence versus no presence (0.79 [0.62-0.99]). The time period (proxy variable for the development of endoscopes) had no impact on completion rates.
Conclusions: The completion rate was lower than that has been reported from previous hospital-based studies. Lower completion rates were found in women, older patients and in patients with diverticulosis. Higher completion rates were noted in patients with previous colonic surgery and long-standing colitis. The completion rates were not influenced by the development of the endoscopes.