Complete colonoscopy: how often? And if not, why not?

OBJECTIVE: Colonoscopy completion rate is an easily measurable criterion of technical competency. Reporting of completion rates lacks uniformity, however, and few studies focus on colonoscopy completion alone. The purpose of this study is to establish criteria for consistency in the reporting of completion rates, so that colonoscopists are better able to use such reports to evaluate their own experience.
METHODS: A prospective study of colonoscopy completion rate and reasons for incompletion was carried out for 2907 patients. Completion was defined as the colonoscope touching the end of the colon. Rates are reported as crude (all cases) and adjusted (excluding incompletions due to stool and disease).
RESULTS: The crude completion rate was 93.6% and the adjusted rate was 98.8%. Reasons for incompletion were stool (n = 47), colonic disease (n = 97), and pain or tortuosity (n = 34). The crude completion rate was lower in women than in men (92.4% vs. 94.8%), lower in the very young (< 20 yr, 85.7%) and very old (> 80 yr, 88.9%), was < 90% in patients presenting with altered bowel habit, diarrhea, constipation, hemorrhage, inflammatory bowel disease, abdominal pain, or cancer, was only 53.8% in patients in intensive care units, was 84.1% in the author's first 127 cases, was lower in women post hysterectomy (92.8% vs. 98.3%), and was higher in patients who had had a colon resection [98.4% (right colectomy), 99.2% (left colectomy), 95.8% (intact colon)]. When adjusted rates were compared, most of those differences disappeared (except male vs. female, hysterectomy vs. no hysterectomy).
CONCLUSIONS: Crude colonoscopy completion rates are affected by a number of factors that may make comparisons between colonoscopists difficult. The use of adjusted completion rates minimizes the effect of disease-related factors, allows completion rate to be a better reflection of technical ability, and may facilitate more uniform reporting of colonoscopy results.