Assessing bowel preparation quality using the mean number of adenomas per colonoscopy.

INTRODUCTION The quality of the bowel preparation directly influences colonoscopy effectiveness. Quality indicators are widely employed to monitor operator performance and to gauge colonoscopy effectiveness. Some have suggested that the enumeration of the mean number of adenomas per colonoscopy (MNA) may be a more useful measure of bowel preparation quality, but evidence of the utility of this metric is limited. The relationship between bowel preparation quality and MNA was assessed. METHODS Records of adult patients, aged 50-74 years, who had undergone a screening colonoscopy in a 6 month period at a hospital-based endoscopy suite in New York City were examined. Excluded were those who were symptomatic or having a colonoscopy for surveillance. Patient and procedural characteristics and clinical findings were abstracted from the endoscopy database. Bowel preparation quality was recorded as excellent, good, fair and poor. Histology and size of polyps removed were gathered from pathology reports. MNA was calculated and incident rate ratios assessing the relationship between bowel preparation quality, MNA, and covariates was calculated using Poisson regression. RESULTS A total of 2422 colonoscopies were identified; 815 (33.6%) were screening colonoscopies among average risk individuals, 50-74 years; 203 (24.9%) had ≥1 adenomas; and 666 (81.7%) had excellent/good preparation quality. Overall MNA was 0.34 [standard deviation (SD) 0.68] and MNA was greater among those >60 years [incident rate ratio (IRR) 1.89, 95% confidence interval (CI) 1.48-2.42), males (IRR 1.60, 95%CI 1.26-2.04) and those with good bowel preparation (IRR 2.54, 95%CI 1.04-6.16). Among those with ≥1 adenomas, MNA was 1.48 (SD 1.05) for excellent and 1.00 (SD 0.00) for poor quality preparation (p = 0.55). CONCLUSIONS We found that MNA is sensitive to changes in bowel preparation with higher MNA among those with good bowel preparation compared with those with poor preparation. Our evidence suggests MNA was particularly sensitive when restricted to only those in whom adenomas were seen.